Apple-2-Tools/ciderpress
1
0
Fork 0
mirror of https://github.com/fadden/ciderpress.git synced 2024-11-26 17:49:21 +00:00
Code Issues Projects Releases Wiki Activity
508475a094
ciderpress/app/DiskArchive.cpp

2903 lines
96 KiB
C++
Raw Normal View History

Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* CiderPress
* Copyright (C) 2009 by CiderPress authors. All Rights Reserved.
* Copyright (C) 2007 by faddenSoft, LLC. All Rights Reserved.
* See the file LICENSE for distribution terms.
*/
/*
* Bridge between DiskImg and GenericArchive.
*/
#include "stdafx.h"
#include "DiskArchive.h"
#include "NufxArchive.h"
#include "Preferences.h"
#include "Main.h"
#include "ImageFormatDialog.h"
#include "RenameEntryDialog.h"
#include "ConfirmOverwriteDialog.h"
#include "../diskimg/DiskImgDetail.h"
Fix subvol charset handling We weren't doing a MOR-to-UNI conversion on the sub-volume name, so HFS volumes with non-ASCII characters didn't look right. This also relocates the character-conversion code to a new source file. It's currently part of the reformat lib, though it arguably belongs in util (but that would introduce a new dependency between reformat and util).
2015-01-15 18:22:08 +00:00
#include "../reformat/Charset.h"
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
static const char kEmptyFolderMarker[] = ".$$EmptyFolder";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* ===========================================================================
* DiskEntry
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
int DiskEntry::ExtractThreadToBuffer(int which, char** ppText, long* pLength,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
CString* pErrMsg) const
{
DIError dierr;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
A2FileDescr* pOpenFile = NULL;
char* dataBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
bool rsrcFork;
bool needAlloc = true;
int result = -1;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(fpFile != NULL);
ASSERT(pErrMsg != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
*pErrMsg = "";
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (*ppText != NULL)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
needAlloc = false;
if (GetDamaged()) {
*pErrMsg = "File is damaged";
goto bail;
}
if (which == kDataThread)
rsrcFork = false;
else if (which == kRsrcThread)
rsrcFork = true;
else {
*pErrMsg = "No such fork";
goto bail;
}
LONGLONG len;
if (rsrcFork)
len = fpFile->GetRsrcLength();
else
len = fpFile->GetDataLength();
if (len == 0) {
if (needAlloc) {
*ppText = new char[1];
**ppText = '\0';
}
*pLength = 0;
result = IDOK;
goto bail;
} else if (len < 0) {
assert(rsrcFork); // forked files always have a data fork
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
*pErrMsg = L"That fork doesn't exist";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
dierr = fpFile->Open(&pOpenFile, true, rsrcFork);
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
*pErrMsg = L"File open failed";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
SET_PROGRESS_BEGIN();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
pOpenFile->SetProgressUpdater(DiskArchive::ProgressCallback, len, NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (needAlloc) {
dataBuf = new char[(int) len];
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (dataBuf == NULL) {
Fix some issues identified by static analyzer Includes "printf format" annotation on debug log function.
2014-11-19 22:54:24 +00:00
pErrMsg->Format(L"ERROR: allocation of %I64d bytes failed", len);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
} else {
if (*pLength < (long) len) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pErrMsg->Format(L"ERROR: buf size %ld too short (%ld)",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
*pLength, (long) len);
goto bail;
}
dataBuf = *ppText;
}
dierr = pOpenFile->Read(dataBuf, (size_t) len);
if (dierr != kDIErrNone) {
if (dierr == kDIErrCancelled) {
result = IDCANCEL;
} else {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pErrMsg->Format(L"File read failed: %hs",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
}
goto bail;
}
if (needAlloc)
*ppText = dataBuf;
*pLength = (long) len;
result = IDOK;
bail:
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pOpenFile != NULL)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pOpenFile->Close();
if (result == IDOK) {
SET_PROGRESS_END();
ASSERT(pErrMsg->IsEmpty());
} else {
ASSERT(result == IDCANCEL || !pErrMsg->IsEmpty());
if (needAlloc) {
delete[] dataBuf;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(*ppText == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
}
return result;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
int DiskEntry::ExtractThreadToFile(int which, FILE* outfp, ConvertEOL conv,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ConvertHighASCII convHA, CString* pErrMsg) const
{
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
A2FileDescr* pOpenFile = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
bool rsrcFork;
int result = -1;
ASSERT(IDOK != -1 && IDCANCEL != -1);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(fpFile != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (which == kDataThread)
rsrcFork = false;
else if (which == kRsrcThread)
rsrcFork = true;
else {
/* if we handle disk images, make sure we disable "conv" */
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
*pErrMsg = L"No such fork";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
LONGLONG len;
if (rsrcFork)
len = fpFile->GetRsrcLength();
else
len = fpFile->GetDataLength();
if (len == 0) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Empty fork");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
result = IDOK;
goto bail;
} else if (len < 0) {
assert(rsrcFork); // forked files always have a data fork
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
*pErrMsg = L"That fork doesn't exist";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
DIError dierr;
dierr = fpFile->Open(&pOpenFile, true, rsrcFork);
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
*pErrMsg = L"Unable to open file on disk image";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
dierr = CopyData(pOpenFile, outfp, conv, convHA, pErrMsg);
if (dierr != kDIErrNone) {
if (pErrMsg->IsEmpty()) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pErrMsg->Format(L"Failed while copying data: %hs\n",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
}
goto bail;
}
result = IDOK;
bail:
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pOpenFile != NULL)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pOpenFile->Close();
return result;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
DIError DiskEntry::CopyData(A2FileDescr* pOpenFile, FILE* outfp, ConvertEOL conv,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ConvertHighASCII convHA, CString* pMsg) const
{
DIError dierr = kDIErrNone;
const int kChunkSize = 16384;
char buf[kChunkSize];
//bool firstChunk = true;
//EOLType sourceType;
bool lastCR = false;
LONGLONG srcLen, dataRem;
/* get the length of the open file */
dierr = pOpenFile->Seek(0, DiskImgLib::kSeekEnd);
if (dierr != kDIErrNone)
goto bail;
srcLen = pOpenFile->Tell();
dierr = pOpenFile->Rewind();
if (dierr != kDIErrNone)
goto bail;
ASSERT(srcLen > 0); // empty files should've been caught earlier
SET_PROGRESS_BEGIN();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
pOpenFile->SetProgressUpdater(DiskArchive::ProgressCallback, srcLen, NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* Loop until all data copied.
*/
dataRem = srcLen;
while (dataRem) {
int chunkLen;
if (dataRem > kChunkSize)
chunkLen = kChunkSize;
else
chunkLen = (int) dataRem;
/* read a chunk from the source file */
dierr = pOpenFile->Read(buf, chunkLen);
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pMsg->Format(L"File read failed: %hs",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
goto bail;
}
/* write chunk to destination file */
int err = GenericEntry::WriteConvert(outfp, buf, chunkLen, &conv,
&convHA, &lastCR);
if (err != 0) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pMsg->Format(L"File write failed: %hs", strerror(err));
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dierr = kDIErrGeneric;
goto bail;
}
dataRem -= chunkLen;
//SET_PROGRESS_UPDATE(ComputePercent(srcLen - dataRem, srcLen));
}
bail:
pOpenFile->ClearProgressUpdater();
SET_PROGRESS_END();
return dierr;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskEntry::GetFeatureFlag(Feature feature) const
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
DiskImg::FSFormat format;
format = fpFile->GetDiskFS()->GetDiskImg()->GetFSFormat();
switch (feature) {
case kFeatureCanChangeType:
{
//if (GetRecordKind() == kRecordKindVolumeDir)
// return false;
switch (format) {
case DiskImg::kFormatProDOS:
case DiskImg::kFormatPascal:
case DiskImg::kFormatMacHFS:
case DiskImg::kFormatDOS32:
case DiskImg::kFormatDOS33:
return true;
default:
return false;
}
}
case kFeaturePascalTypes:
{
switch (format) {
case DiskImg::kFormatPascal:
return true;
default:
return false;
}
}
case kFeatureDOSTypes:
{
switch (format) {
case DiskImg::kFormatDOS32:
case DiskImg::kFormatDOS33:
return true;
default:
return false;
}
}
case kFeatureHFSTypes:
{
switch (format) {
case DiskImg::kFormatMacHFS:
return true;
default:
return false;
}
}
case kFeatureHasFullAccess:
{
switch (format) {
case DiskImg::kFormatProDOS:
return true;
default:
return false;
}
}
case kFeatureHasSimpleAccess:
{
switch (format) {
case DiskImg::kFormatDOS33:
case DiskImg::kFormatDOS32:
case DiskImg::kFormatCPM:
case DiskImg::kFormatMacHFS:
return true;
default:
return false;
}
}
case kFeatureHasInvisibleFlag:
{
switch(format) {
case DiskImg::kFormatProDOS:
case DiskImg::kFormatMacHFS:
return true;
default:
return false;
}
}
default:
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Unexpected feature flag %d", feature);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
assert(false);
return false;
}
assert(false);
return false;
}
/*
* ===========================================================================
* DiskArchive
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*static*/ CString DiskArchive::AppInit(void)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
CString result("");
DIError dierr;
Update NufxLib snapshot to v3.0.0d2 This integrates the latest NufxLib sources, and updates CiderPress to work with the API changes.
2015-01-04 19:29:51 +00:00
int32_t major, minor, bug;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Initializing DiskImg library");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
// set this before initializing, so we can see init debug msgs
DiskImgLib::Global::SetDebugMsgHandler(DebugMsgHandler);
dierr = DiskImgLib::Global::AppInit();
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
result.Format(L"DiskImg DLL failed to initialize: %hs\n",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
goto bail;
}
DiskImgLib::Global::GetVersion(&major, &minor, &bug);
if (major != kDiskImgVersionMajor || minor < kDiskImgVersionMinor) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
result.Format(L"Older or incompatible version of DiskImg DLL found.\r\r"
L"Wanted v%d.%d.x, found %ld.%ld.%ld.",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
kDiskImgVersionMajor, kDiskImgVersionMinor,
major, minor, bug);
goto bail;
}
bail:
return result;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*static*/ void DiskArchive::AppCleanup(void)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
DiskImgLib::Global::AppCleanup();
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*static*/ void DiskArchive::DebugMsgHandler(const char* file, int line,
const char* msg)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(file != NULL);
ASSERT(msg != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Added new logging macros Visual Studio figured out variadic macros around 2005, so we can finally replace the explicit-arg-count debug log macros. Also, fixed some include guards. Also, bumped version to 4.0.0d1.
2014-11-18 01:54:34 +00:00
LOG_BASE(DebugLog::LOG_INFO, file, line, "<diskimg> %hs", msg);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*static*/ bool DiskArchive::ProgressCallback(DiskImgLib::A2FileDescr* pFile,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::di_off_t max, DiskImgLib::di_off_t current, void* state)
{
int status;
//::Sleep(10);
status = SET_PROGRESS_UPDATE(ComputePercent(current, max));
if (status == IDCANCEL) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("IDCANCEL returned from Main progress updater");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
return false;
}
return true; // tell DiskImgLib to continue what it's doing
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*static*/ bool DiskArchive::ScanProgressCallback(void* cookie, const char* str,
int count)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
CString fmt;
bool cont;
if (count == 0)
fmt = str;
else
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
fmt.Format(L"%hs (%%d)", str);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
cont = SET_PROGRESS_COUNTER_2(fmt, count);
if (!cont) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("cancelled");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
return cont;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
GenericArchive::OpenResult DiskArchive::Open(const WCHAR* filename,
bool readOnly, CString* pErrMsg)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
DIError dierr;
CString errMsg;
OpenResult result = kResultUnknown;
const Preferences* pPreferences = GET_PREFERENCES();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(fpPrimaryDiskFS == NULL);
ASSERT(filename != NULL);
//ASSERT(ext != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pPreferences != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
fIsReadOnly = readOnly;
// special case for volume open
bool isVolume = false;
if (filename[0] >= 'A' && filename[0] <= 'Z' &&
filename[1] == ':' && filename[2] == '\\' &&
filename[3] == '\0')
{
isVolume = true;
}
/*
* Open the image. This can be very slow for compressed images,
* especially 3.5" FDI images.
*/
{
CWaitCursor waitc;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
// TODO(Unicode): modify DiskImg lib to accept wide paths
Fix diskimg OpenImage The OpenImage method had an overload that took void*. This turns out to be a bad idea, because void* matches any pointer type that didn't match something else. So the WCHAR* filenames were going to the "open from buffer" method rather than the "open from file" variant. A less important issue is whether open-from-buffer should take a const or non-const pointer. If the "readOnly" boolean flag is not set, then the contents can be altered and const is inappropriate. The best course seems to be to drop the boolean flag as an argument, and just have two different methods.
2014-11-23 18:34:57 +00:00
CStringA fileNameA(filename);
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
if (!PathName::TestNarrowConversion(filename, fileNameA, &errMsg)) {
result = kResultFailure;
goto bail;
}
Fix diskimg OpenImage The OpenImage method had an overload that took void*. This turns out to be a bad idea, because void* matches any pointer type that didn't match something else. So the WCHAR* filenames were going to the "open from buffer" method rather than the "open from file" variant. A less important issue is whether open-from-buffer should take a const or non-const pointer. If the "readOnly" boolean flag is not set, then the contents can be altered and const is inappropriate. The best course seems to be to drop the boolean flag as an argument, and just have two different methods.
2014-11-23 18:34:57 +00:00
dierr = fDiskImg.OpenImage(fileNameA, PathProposal::kLocalFssep,
readOnly);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (dierr == kDIErrAccessDenied && !readOnly && !isVolume) {
// retry file open with read-only set
// don't do that for volumes -- assume they know what they want
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD(" Retrying open with read-only set");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
fIsReadOnly = readOnly = true;
Fix diskimg OpenImage The OpenImage method had an overload that took void*. This turns out to be a bad idea, because void* matches any pointer type that didn't match something else. So the WCHAR* filenames were going to the "open from buffer" method rather than the "open from file" variant. A less important issue is whether open-from-buffer should take a const or non-const pointer. If the "readOnly" boolean flag is not set, then the contents can be altered and const is inappropriate. The best course seems to be to drop the boolean flag as an argument, and just have two different methods.
2014-11-23 18:34:57 +00:00
dierr = fDiskImg.OpenImage(fileNameA, PathProposal::kLocalFssep,
readOnly);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
if (dierr != kDIErrNone) {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
if (dierr == kDIErrFileArchive) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
result = kResultFileArchive;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
} else {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
result = kResultFailure;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to open '%ls': %hs.", filename,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
}
goto bail;
}
}
dierr = fDiskImg.AnalyzeImage();
if (dierr != kDIErrNone) {
result = kResultFailure;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Analysis of '%ls' failed: %hs", filename,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
goto bail;
}
/* allow them to override sector order and filesystem, if requested */
if (pPreferences->GetPrefBool(kPrQueryImageFormat)) {
ImageFormatDialog imf;
imf.InitializeValues(&fDiskImg);
imf.fFileSource = filename;
imf.SetQueryDisplayFormat(false);
imf.SetAllowGenericFormats(false);
if (imf.DoModal() != IDOK) {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD("User bailed on IMF dialog");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
result = kResultCancel;
goto bail;
}
if (imf.fSectorOrder != fDiskImg.GetSectorOrder() ||
imf.fFSFormat != fDiskImg.GetFSFormat())
{
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Initial values overridden, forcing img format");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dierr = fDiskImg.OverrideFormat(fDiskImg.GetPhysicalFormat(),
imf.fFSFormat, imf.fSectorOrder);
if (dierr != kDIErrNone) {
result = kResultFailure;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to access disk image using selected"
L" parameters. Error: %hs.",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
goto bail;
}
}
}
if (fDiskImg.GetFSFormat() == DiskImg::kFormatUnknown ||
fDiskImg.GetSectorOrder() == DiskImg::kSectorOrderUnknown)
{
result = kResultFailure;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to identify filesystem on '%ls'", filename);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
/* create an appropriate DiskFS object */
fpPrimaryDiskFS = fDiskImg.OpenAppropriateDiskFS();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (fpPrimaryDiskFS == NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/* unknown FS should've been caught above! */
ASSERT(false);
result = kResultFailure;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Format of '%ls' not recognized.", filename);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
fpPrimaryDiskFS->SetScanForSubVolumes(DiskFS::kScanSubEnabled);
/*
* Scan all files and on the disk image, and recursively descend into
* sub-volumes. Can be slow on physical volumes.
*
* This is really only useful for ProDOS and HFS disks. Nothing else
* can be large enough to really get slow, and nothing else is likely
* to show up in a large multi-partition image.
*
* THOUGHT: only show the dialog if the volume is over a certain size.
*/
{
MainWindow* pMain = GET_MAIN_WINDOW();
ProgressCounterDialog* pProgress;
pProgress = new ProgressCounterDialog;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pProgress->Create(L"Examining contents, please wait...", pMain);
pProgress->SetCounterFormat(L"Scanning...");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pProgress->CenterWindow();
//pMain->PeekAndPump(); // redraw
CWaitCursor waitc;
/* set up progress dialog and scan all files */
pMain->SetProgressCounterDialog(pProgress);
fDiskImg.SetScanProgressCallback(ScanProgressCallback, this);
dierr = fpPrimaryDiskFS->Initialize(&fDiskImg, DiskFS::kInitFull);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
fDiskImg.SetScanProgressCallback(NULL, NULL);
pMain->SetProgressCounterDialog(NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pProgress->DestroyWindow();
if (dierr != kDIErrNone) {
if (dierr == kDIErrCancelled) {
result = kResultCancel;
} else {
result = kResultFailure;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Error reading list of files from disk: %hs",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
}
goto bail;
}
}
if (LoadContents() != 0) {
result = kResultFailure;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg = L"Failed while loading contents of disk image.";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
/*
* Force read-only flag if underlying FS doesn't allow RW. We need to
* consider embedded filesystems, so we only set RO if none of the
* filesystems are writable.
*
* BUG: this only checks the first level. Should be fully recursive.
*/
if (!fpPrimaryDiskFS->GetReadWriteSupported()) {
const DiskFS::SubVolume* pSubVol;
fIsReadOnly = true;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
pSubVol = fpPrimaryDiskFS->GetNextSubVolume(NULL);
while (pSubVol != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (pSubVol->GetDiskFS()->GetReadWriteSupported()) {
fIsReadOnly = false;
break;
}
pSubVol = fpPrimaryDiskFS->GetNextSubVolume(pSubVol);
}
}
/* force read-only if the primary is damaged */
if (fpPrimaryDiskFS->GetFSDamaged())
fIsReadOnly = true;
/* force read-only if the DiskImg thinks a wrapper is damaged */
if (fpPrimaryDiskFS->GetDiskImg()->GetReadOnly())
fIsReadOnly = true;
// /* force read-only on .gz/.zip unless pref allows */
// if (fDiskImg.GetOuterFormat() != DiskImg::kOuterFormatNone) {
// if (pPreferences->GetPrefBool(kPrWriteZips) == 0)
// fIsReadOnly = true;
// }
SetPathName(filename);
result = kResultSuccess;
/* set any preferences-based settings */
PreferencesChanged();
bail:
*pErrMsg = errMsg;
if (!errMsg.IsEmpty()) {
assert(result == kResultFailure);
delete fpPrimaryDiskFS;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
fpPrimaryDiskFS = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
} else {
assert(result != kResultFailure);
}
return result;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
CString DiskArchive::New(const WCHAR* fileName, const void* vOptions)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
const Preferences* pPreferences = GET_PREFERENCES();
NewOptions* pOptions = (NewOptions*) vOptions;
CString volName;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
CStringA volNameA, fileNameA;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
long numBlocks = -1;
long numTracks = -1;
int numSectors;
CString retmsg;
DIError dierr;
bool allowLowerCase;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(fileName != NULL);
ASSERT(pOptions != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
allowLowerCase = pPreferences->GetPrefBool(kPrProDOSAllowLower) != 0;
switch (pOptions->base.format) {
case DiskImg::kFormatUnknown:
numBlocks = pOptions->blank.numBlocks;
break;
case DiskImg::kFormatProDOS:
volName = pOptions->prodos.volName;
numBlocks = pOptions->prodos.numBlocks;
break;
case DiskImg::kFormatPascal:
volName = pOptions->pascalfs.volName;
numBlocks = pOptions->pascalfs.numBlocks;
break;
case DiskImg::kFormatMacHFS:
volName = pOptions->hfs.volName;
numBlocks = pOptions->hfs.numBlocks;
break;
case DiskImg::kFormatDOS32:
numTracks = pOptions->dos.numTracks;
numSectors = pOptions->dos.numSectors;
if (numTracks < DiskFSDOS33::kMinTracks ||
numTracks > DiskFSDOS33::kMaxTracks)
{
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg = L"Invalid DOS32 track count";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
if (numSectors != 13) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg = L"Invalid DOS32 sector count";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
if (pOptions->dos.allocDOSTracks)
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
volName = L"DOS";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatDOS33:
numTracks = pOptions->dos.numTracks;
numSectors = pOptions->dos.numSectors;
if (numTracks < DiskFSDOS33::kMinTracks ||
numTracks > DiskFSDOS33::kMaxTracks)
{
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg = L"Invalid DOS33 track count";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
if (numSectors != 16 && numSectors != 32) { // no 13-sector (yet)
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg = L"Invalid DOS33 sector count";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
if (pOptions->dos.allocDOSTracks)
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
volName = L"DOS";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
default:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg = L"Unsupported disk format";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("DiskArchive: new '%ls' %ld %hs in '%ls'",
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
(LPCWSTR) volName, numBlocks,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImg::ToString(pOptions->base.format), fileName);
bool canSkipFormat;
if (IsWin9x())
canSkipFormat = false;
else
canSkipFormat = true;
/*
* Create an image with the appropriate characteristics. We set
* "skipFormat" because we know this will be a brand-new file, and
* we're not currently creating nibble images.
*
* GLITCH: under Win98/ME, brand-new files contain the previous contents
* of the hard drive. We need to explicitly zero them out. We don't
* want to do it under Win2K/XP because it can be slow for larger
* volumes.
*/
Fix subvol charset handling We weren't doing a MOR-to-UNI conversion on the sub-volume name, so HFS volumes with non-ASCII characters didn't look right. This also relocates the character-conversion code to a new source file. It's currently part of the reformat lib, though it arguably belongs in util (but that would introduce a new dependency between reformat and util).
2015-01-15 18:22:08 +00:00
fileNameA = fileName; // TODO(Unicode)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (numBlocks > 0) {
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
dierr = fDiskImg.CreateImage(fileNameA, NULL,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImg::kOuterFormatNone,
DiskImg::kFileFormatUnadorned,
DiskImg::kPhysicalFormatSectors,
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
NULL,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pOptions->base.sectorOrder,
DiskImg::kFormatGenericProDOSOrd, // arg must be generic
numBlocks,
canSkipFormat);
} else {
ASSERT(numTracks > 0);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
dierr = fDiskImg.CreateImage(fileNameA, NULL,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImg::kOuterFormatNone,
DiskImg::kFileFormatUnadorned,
DiskImg::kPhysicalFormatSectors,
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
NULL,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pOptions->base.sectorOrder,
DiskImg::kFormatGenericProDOSOrd, // arg must be generic
numTracks, numSectors,
canSkipFormat);
}
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg.Format(L"Unable to create disk image: %hs.",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
goto bail;
}
if (pOptions->base.format == DiskImg::kFormatUnknown)
goto skip_format;
if (pOptions->base.format == DiskImg::kFormatDOS33 ||
pOptions->base.format == DiskImg::kFormatDOS32)
fDiskImg.SetDOSVolumeNum(pOptions->dos.volumeNum);
/*
* If we don't allow lower case in ProDOS filenames, don't allow them
* in volume names either. This works because we don't allow ' ' in
* volume names; otherwise we'd need to invoke a ProDOS-specific call
* to convert the ' ' to '.'. (Or we could just do it ourselves.)
*
* We can't ask the ProDOS DiskFS to force upper case for us because
* the ProDOS DiskFS object doesn't yet exist.
*/
if (pOptions->base.format == DiskImg::kFormatProDOS && !allowLowerCase)
volName.MakeUpper();
/* format it */
Fix subvol charset handling We weren't doing a MOR-to-UNI conversion on the sub-volume name, so HFS volumes with non-ASCII characters didn't look right. This also relocates the character-conversion code to a new source file. It's currently part of the reformat lib, though it arguably belongs in util (but that would introduce a new dependency between reformat and util).
2015-01-15 18:22:08 +00:00
// TODO(Unicode): for HFS, we need to convert Unicode to MOR
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
volNameA = volName;
dierr = fDiskImg.FormatImage(pOptions->base.format, volNameA);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg.Format(L"Unable to format disk image: %hs.",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
goto bail;
}
fpPrimaryDiskFS = fDiskImg.OpenAppropriateDiskFS(false);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (fpPrimaryDiskFS == NULL) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg = L"Unable to create DiskFS.";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
/* prep it */
dierr = fpPrimaryDiskFS->Initialize(&fDiskImg, DiskFS::kInitFull);
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg.Format(L"Error reading list of files from disk: %hs",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
goto bail;
}
/* this is pretty meaningless, but do it to ensure we're initialized */
if (LoadContents() != 0) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
retmsg = L"Failed while loading contents of disk image.";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
skip_format:
SetPathName(fileName);
/* set any preferences-based settings */
PreferencesChanged();
bail:
return retmsg;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
CString DiskArchive::Close(void)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (fpPrimaryDiskFS != NULL) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("DiskArchive shutdown closing disk image");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
delete fpPrimaryDiskFS;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
fpPrimaryDiskFS = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
DIError dierr;
dierr = fDiskImg.CloseImage();
if (dierr != kDIErrNone) {
MainWindow* pMainWin = (MainWindow*)::AfxGetMainWnd();
CString msg, failed;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
msg.Format(L"Failed while closing disk image: %hs.",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&failed, IDS_FAILED);
Fix some issues identified by static analyzer Includes "printf format" annotation on debug log function.
2014-11-19 22:54:24 +00:00
LOGE("During close: %ls", (LPCWSTR) msg);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pMainWin->MessageBox(msg, failed, MB_OK);
}
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
return L"";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
CString DiskArchive::Flush(void)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
DIError dierr;
CWaitCursor waitc;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
assert(fpPrimaryDiskFS != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dierr = fpPrimaryDiskFS->Flush(DiskImg::kFlushAll);
if (dierr != kDIErrNone) {
CString errMsg;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Attempt to flush the current archive failed: %hs.",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
return errMsg;
}
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
return L"";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskArchive::IsModified(void) const
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
assert(fpPrimaryDiskFS != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
return fpPrimaryDiskFS->GetDiskImg()->GetDirtyFlag();
}
Rewrote "open archive" logic The previous code was a stickler for only opening files whose type matched what was selected in the filter pop-up. The original goal was to allow you to choose whether a BXY or SDK file was interpreted as Binary II, ShrinkIt, or disk image, since they could go either way. Unfortunately, its refusal to consider types other than what was selected made it kind of annoying. The new code will start by trying to open the file with the selected filter, so that it's still possible to choose how SDK and BXY files are opened. However, it now continues on, trying all other types before finally giving up. If the generic ("*.*") filter is selected, CiderPress will start by trying to open the file as a disk image. This seems to produce good results with a variety of known and unknown files.
2015-01-13 23:54:10 +00:00
CString DiskArchive::GetDescription() const
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Rewrote "open archive" logic The previous code was a stickler for only opening files whose type matched what was selected in the filter pop-up. The original goal was to allow you to choose whether a BXY or SDK file was interpreted as Binary II, ShrinkIt, or disk image, since they could go either way. Unfortunately, its refusal to consider types other than what was selected made it kind of annoying. The new code will start by trying to open the file with the selected filter, so that it's still possible to choose how SDK and BXY files are opened. However, it now continues on, trying all other types before finally giving up. If the generic ("*.*") filter is selected, CiderPress will start by trying to open the file as a disk image. This seems to produce good results with a variety of known and unknown files.
2015-01-13 23:54:10 +00:00
CString str = L"Disk Image";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Rewrote "open archive" logic The previous code was a stickler for only opening files whose type matched what was selected in the filter pop-up. The original goal was to allow you to choose whether a BXY or SDK file was interpreted as Binary II, ShrinkIt, or disk image, since they could go either way. Unfortunately, its refusal to consider types other than what was selected made it kind of annoying. The new code will start by trying to open the file with the selected filter, so that it's still possible to choose how SDK and BXY files are opened. However, it now continues on, trying all other types before finally giving up. If the generic ("*.*") filter is selected, CiderPress will start by trying to open the file as a disk image. This seems to produce good results with a variety of known and unknown files.
2015-01-13 23:54:10 +00:00
if (fpPrimaryDiskFS != NULL && fpPrimaryDiskFS->GetVolumeID() != NULL) {
Simplify
2015-01-15 22:21:01 +00:00
CString volumeId(Charset::ConvertMORToUNI(fpPrimaryDiskFS->GetVolumeID()));
Volume name and font fixes More volume name MOR conversions. I think I got them all. This also switches the "archive info", "add files", and "extract files" dialogs to use the System Font. We were using "MS Sans Serif" before, which looks a bit ratty on Windows 7 because it doesn't take advantage of ClearType. (Apparently the ClearType version is "Microsoft Sans Serif", though when you set the "use system font" boolean to true it changes the font name to "MS Shell Dlg".) The old font also seems to be missing certain glyphs, e.g. my HFS volume name had 'TM' in it, but that just showed up as a box (which is why, in case you were wondering, these changes ended up together). The new font seems to work equally well on WinXP, so I may enable it for all dialogs in a follow-up change. As far as I can tell it has the same font metrics -- I haven't seen anything weird looking in the dialogs I've updated so far. Also, bumped the version to 4.0.0-b3.
2015-01-15 19:25:26 +00:00
str.Format(L"Disk Image - %ls", (LPCWSTR) volumeId);
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
}
Rewrote "open archive" logic The previous code was a stickler for only opening files whose type matched what was selected in the filter pop-up. The original goal was to allow you to choose whether a BXY or SDK file was interpreted as Binary II, ShrinkIt, or disk image, since they could go either way. Unfortunately, its refusal to consider types other than what was selected made it kind of annoying. The new code will start by trying to open the file with the selected filter, so that it's still possible to choose how SDK and BXY files are opened. However, it now continues on, trying all other types before finally giving up. If the generic ("*.*") filter is selected, CiderPress will start by trying to open the file as a disk image. This seems to produce good results with a variety of known and unknown files.
2015-01-13 23:54:10 +00:00
return str;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
int DiskArchive::LoadContents(void)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
int result;
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("DiskArchive LoadContents");
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(fpPrimaryDiskFS != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
MainWindow* pMain = GET_MAIN_WINDOW();
ExclusiveModelessDialog* pWaitDlg = new ExclusiveModelessDialog;
pWaitDlg->Create(IDD_LOADING, pMain);
pWaitDlg->CenterWindow();
pMain->PeekAndPump(); // redraw
CWaitCursor waitc;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
result = LoadDiskFSContents(fpPrimaryDiskFS, L"");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
SET_PROGRESS_COUNTER(-1);
pWaitDlg->DestroyWindow();
//pMain->PeekAndPump(); // redraw
}
return result;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
CString DiskArchive::Reload()
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
fReloadFlag = true; // tell everybody that cached data is invalid
(void) fpPrimaryDiskFS->Flush(DiskImg::kFlushFastOnly);
DeleteEntries(); // a GenericArchive operation
if (LoadContents() != 0)
return "Disk image reload failed.";
return "";
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
int DiskArchive::InternalReload(CWnd* pMsgWnd)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
CString errMsg;
errMsg = Reload();
if (!errMsg.IsEmpty()) {
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
return -1;
}
return 0;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
int DiskArchive::LoadDiskFSContents(DiskFS* pDiskFS, const WCHAR* volName)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
static const char* kBlankFileNameMOR = "<blank filename>";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
A2File* pFile;
DiskEntry* pNewEntry;
DiskFS::SubVolume* pSubVol;
const Preferences* pPreferences = GET_PREFERENCES();
bool wantCoerceDOSFilenames = false;
CString ourSubVolName;
wantCoerceDOSFilenames = pPreferences->GetPrefBool(kPrCoerceDOSFilenames);
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Notes for disk image '%ls':", volName);
LOGI("%hs", pDiskFS->GetDiskImg()->GetNotes());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pDiskFS != NULL);
pFile = pDiskFS->GetNextFile(NULL);
while (pFile != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pNewEntry = new DiskEntry(pFile);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pNewEntry == NULL)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
return -1;
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
CStringA path(pFile->GetPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (path.IsEmpty())
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
path = kBlankFileNameMOR;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (DiskImg::UsesDOSFileStructure(pFile->GetFSFormat()) &&
wantCoerceDOSFilenames)
{
InjectLowercase(&path);
}
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
pNewEntry->SetPathNameMOR(path);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (volName[0] != '\0')
pNewEntry->SetSubVolName(volName);
pNewEntry->SetFssep(pFile->GetFssep());
pNewEntry->SetFileType(pFile->GetFileType());
pNewEntry->SetAuxType(pFile->GetAuxType());
pNewEntry->SetAccess(pFile->GetAccess());
if (pFile->GetCreateWhen() == 0)
pNewEntry->SetCreateWhen(kDateNone);
else
pNewEntry->SetCreateWhen(pFile->GetCreateWhen());
if (pFile->GetModWhen() == 0)
pNewEntry->SetModWhen(kDateNone);
else
pNewEntry->SetModWhen(pFile->GetModWhen());
pNewEntry->SetSourceFS(pFile->GetFSFormat());
pNewEntry->SetHasDataFork(true);
if (pFile->IsVolumeDirectory()) {
/* volume directory entry; only on ProDOS/HFS */
ASSERT(pFile->GetRsrcLength() < 0);
pNewEntry->SetRecordKind(GenericEntry::kRecordKindVolumeDir);
//pNewEntry->SetUncompressedLen(pFile->GetDataLength());
pNewEntry->SetDataForkLen(pFile->GetDataLength());
pNewEntry->SetCompressedLen(pFile->GetDataLength());
} else if (pFile->IsDirectory()) {
/* directory entry */
ASSERT(pFile->GetRsrcLength() < 0);
pNewEntry->SetRecordKind(GenericEntry::kRecordKindDirectory);
//pNewEntry->SetUncompressedLen(pFile->GetDataLength());
pNewEntry->SetDataForkLen(pFile->GetDataLength());
pNewEntry->SetCompressedLen(pFile->GetDataLength());
} else if (pFile->GetRsrcLength() >= 0) {
/* has resource fork */
pNewEntry->SetRecordKind(GenericEntry::kRecordKindForkedFile);
pNewEntry->SetDataForkLen(pFile->GetDataLength());
pNewEntry->SetRsrcForkLen(pFile->GetRsrcLength());
//pNewEntry->SetUncompressedLen(
// pFile->GetDataLength() + pFile->GetRsrcLength() );
pNewEntry->SetCompressedLen(
pFile->GetDataSparseLength() + pFile->GetRsrcSparseLength() );
pNewEntry->SetHasRsrcFork(true);
} else {
/* just data fork */
pNewEntry->SetRecordKind(GenericEntry::kRecordKindFile);
//pNewEntry->SetUncompressedLen(pFile->GetDataLength());
pNewEntry->SetDataForkLen(pFile->GetDataLength());
pNewEntry->SetCompressedLen(pFile->GetDataSparseLength());
}
switch (pNewEntry->GetSourceFS()) {
case DiskImg::kFormatDOS33:
case DiskImg::kFormatDOS32:
case DiskImg::kFormatUNIDOS:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pNewEntry->SetFormatStr(L"DOS");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
Show Gutenberg files as "Gutenb" in the content list
2015-01-10 02:36:05 +00:00
case DiskImg::kFormatGutenberg:
pNewEntry->SetFormatStr(L"Gutenb");
break;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
case DiskImg::kFormatProDOS:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pNewEntry->SetFormatStr(L"ProDOS");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatPascal:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pNewEntry->SetFormatStr(L"Pascal");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatCPM:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pNewEntry->SetFormatStr(L"CP/M");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatMSDOS:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pNewEntry->SetFormatStr(L"MS-DOS");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatRDOS33:
case DiskImg::kFormatRDOS32:
case DiskImg::kFormatRDOS3:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pNewEntry->SetFormatStr(L"RDOS");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatMacHFS:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pNewEntry->SetFormatStr(L"HFS");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
default:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pNewEntry->SetFormatStr(L"???");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
}
pNewEntry->SetDamaged(pFile->GetQuality() == A2File::kQualityDamaged);
pNewEntry->SetSuspicious(pFile->GetQuality() == A2File::kQualitySuspicious);
AddEntry(pNewEntry);
/* this is not very useful -- all the heavy lifting was done earlier */
if ((GetNumEntries() % 100) == 0)
SET_PROGRESS_COUNTER(GetNumEntries());
pFile = pDiskFS->GetNextFile(pFile);
}
/*
* Load all sub-volumes.
*
* We define the sub-volume name to use for the next layer down. We
* prepend an underscore to the unmodified name. So long as the volume
* name is a valid Windows path -- which should hold true for most disks,
* though possibly not for Pascal -- it can be extracted directly with
* its full path with no risk of conflict. (The extraction code relies
* on this, so don't put a ':' in the subvol name or Windows will choke.)
*/
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
pSubVol = pDiskFS->GetNextSubVolume(NULL);
while (pSubVol != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
CString concatSubVolName;
int ret;
Simplify
2015-01-15 22:21:01 +00:00
const char* subVolNameMOR = pSubVol->GetDiskFS()->GetVolumeName();
if (subVolNameMOR == NULL) {
Fix subvol charset handling We weren't doing a MOR-to-UNI conversion on the sub-volume name, so HFS volumes with non-ASCII characters didn't look right. This also relocates the character-conversion code to a new source file. It's currently part of the reformat lib, though it arguably belongs in util (but that would introduce a new dependency between reformat and util).
2015-01-15 18:22:08 +00:00
subVolNameMOR = "+++"; // call it *something*
}
CString subVolName(Charset::ConvertMORToUNI(subVolNameMOR));
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Fix subvol charset handling We weren't doing a MOR-to-UNI conversion on the sub-volume name, so HFS volumes with non-ASCII characters didn't look right. This also relocates the character-conversion code to a new source file. It's currently part of the reformat lib, though it arguably belongs in util (but that would introduce a new dependency between reformat and util).
2015-01-15 18:22:08 +00:00
if (volName[0] == '\0') {
concatSubVolName.Format(L"_%ls", (LPCWSTR) subVolName);
} else {
concatSubVolName.Format(L"%ls_%ls", volName, (LPCWSTR) subVolName);
}
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ret = LoadDiskFSContents(pSubVol->GetDiskFS(), concatSubVolName);
if (ret != 0)
return ret;
pSubVol = pDiskFS->GetNextSubVolume(pSubVol);
}
return 0;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
void DiskArchive::PreferencesChanged(void)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
const Preferences* pPreferences = GET_PREFERENCES();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (fpPrimaryDiskFS != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
fpPrimaryDiskFS->SetParameter(DiskFS::kParmProDOS_AllowLowerCase,
pPreferences->GetPrefBool(kPrProDOSAllowLower) != 0);
fpPrimaryDiskFS->SetParameter(DiskFS::kParmProDOS_AllocSparse,
pPreferences->GetPrefBool(kPrProDOSUseSparse) != 0);
}
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
long DiskArchive::GetCapability(Capability cap)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
switch (cap) {
case kCapCanTest:
return false;
break;
case kCapCanRenameFullPath:
return false;
break;
case kCapCanRecompress:
return false;
break;
case kCapCanEditComment:
return false;
break;
case kCapCanAddDisk:
return false;
break;
case kCapCanConvEOLOnAdd:
return true;
break;
case kCapCanCreateSubdir:
return true;
break;
case kCapCanRenameVolume:
return true;
break;
default:
ASSERT(false);
return -1;
break;
}
}
/*
* ===========================================================================
* DiskArchive -- add files
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskArchive::BulkAdd(ActionProgressDialog* pActionProgress,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
const AddFilesDialog* pAddOpts)
{
NuError nerr;
CString errMsg;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
WCHAR curDir[MAX_PATH] = L"";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
bool retVal = false;
File selection dialog update, part 1 This adds a replacement for the SelectFilesDialog class. It has been updated to use Explorer-style dialogs, which are a bit nicer than the old-style dialogs. Hopefully this will eliminate some of the brain damage, like the disappearing Accept button. This change only updates MDC. A second change will update the main app and remove the old code. Also, updated the MDC version to 3.0.0, and changed the web site linked in the Help menu from faddensoft.com to a2ciderpress.com.
2014-12-02 06:07:39 +00:00
LOGD("Opts: '%ls' typePres=%d", (LPCWSTR) pAddOpts->fStoragePrefix,
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
pAddOpts->fTypePreservation);
File selection dialog update, part 1 This adds a replacement for the SelectFilesDialog class. It has been updated to use Explorer-style dialogs, which are a bit nicer than the old-style dialogs. Hopefully this will eliminate some of the brain damage, like the disappearing Accept button. This change only updates MDC. A second change will update the main app and remove the old code. Also, updated the MDC version to 3.0.0, and changed the web site linked in the Help menu from faddensoft.com to a2ciderpress.com.
2014-12-02 06:07:39 +00:00
LOGD(" sub=%d strip=%d ovwr=%d",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pAddOpts->fIncludeSubfolders, pAddOpts->fStripFolderNames,
pAddOpts->fOverwriteExisting);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(fpAddDataHead == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/* these reset on every new add */
fOverwriteExisting = false;
fOverwriteNoAsk = false;
/* we screen for clashes with existing files later; this just ensures
"batch uniqueness" */
fpPrimaryDiskFS->SetParameter(DiskFS::kParm_CreateUnique, true);
/*
* Save the current directory and change to the one from the file dialog.
*/
File selection dialog update, part 2 This updates the main app to use the new dialog, and removes the old code.
2014-12-04 00:10:03 +00:00
const CString& directory = pAddOpts->GetDirectory();
LOGI("Selected path = '%ls'", (LPCWSTR) directory);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (GetCurrentDirectory(NELEM(curDir), curDir) == 0) {
errMsg = L"Unable to get current directory.\n";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pActionProgress, errMsg, IDS_FAILED);
goto bail;
}
File selection dialog update, part 2 This updates the main app to use the new dialog, and removes the old code.
2014-12-04 00:10:03 +00:00
if (SetCurrentDirectory(directory) == false) {
errMsg.Format(L"Unable to set current directory to '%ls'.\n",
(LPCWSTR) directory);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pActionProgress, errMsg, IDS_FAILED);
goto bail;
}
File selection dialog update, part 2 This updates the main app to use the new dialog, and removes the old code.
2014-12-04 00:10:03 +00:00
const CStringArray& fileNames = pAddOpts->GetFileNames();
for (int i = 0; i < fileNames.GetCount(); i++) {
const CString& name = fileNames.GetAt(i);
LOGI(" file '%ls'", (LPCWSTR) name);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/* add the file, calling DoAddFile via the generic AddFile */
File selection dialog update, part 2 This updates the main app to use the new dialog, and removes the old code.
2014-12-04 00:10:03 +00:00
nerr = AddFile(pAddOpts, name, &errMsg);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (nerr != kNuErrNone) {
if (errMsg.IsEmpty())
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Failed while adding file '%ls': %hs.",
File selection dialog update, part 2 This updates the main app to use the new dialog, and removes the old code.
2014-12-04 00:10:03 +00:00
(LPCWSTR) name, NuStrError(nerr));
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (nerr != kNuErrAborted) {
ShowFailureMsg(pActionProgress, errMsg, IDS_FAILED);
}
goto bail;
}
}
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (fpAddDataHead == NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
CString title;
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&title, IDS_MB_APP_NAME);
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg = L"No files added.\n";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pActionProgress->MessageBox(errMsg, title, MB_OK | MB_ICONWARNING);
} else {
/* add all pending files */
retVal = true;
errMsg = ProcessFileAddData(pAddOpts->fpTargetDiskFS,
pAddOpts->fConvEOL);
if (!errMsg.IsEmpty()) {
CString title;
ShowFailureMsg(pActionProgress, errMsg, IDS_FAILED);
retVal = false;
}
/* success or failure, reload the contents */
errMsg = Reload();
if (!errMsg.IsEmpty())
retVal = false;
}
bail:
FreeAddDataList();
if (SetCurrentDirectory(curDir) == false) {
File selection dialog update, part 2 This updates the main app to use the new dialog, and removes the old code.
2014-12-04 00:10:03 +00:00
errMsg.Format(L"Unable to reset current directory to '%ls'.\n", curDir);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pActionProgress, errMsg, IDS_FAILED);
// bummer, but don't signal failure
}
return retVal;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
NuError DiskArchive::DoAddFile(const AddFilesDialog* pAddOpts,
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LocalFileDetails* pDetails)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*
* Add a file to a disk image Unfortunately we can't just add the files
* here. We need to figure which pairs of files should be combined into
* a single "extended" file (yes, the cursed forked files strike again).
*
* The way you tell if two files should be one is by comparing their
* filenames and type info. If they match, and one is a data fork and
* one is a resource fork, we have a single split file.
*
* We have to be careful here because we don't know which will be seen
* first and whether they'll be adjacent. We have to dig through the
* list of previously-added files for a match (O(n^2) behavior currently).
*
* We also have to compare the right filename. Comparing the Windows
* filename is a bad idea, because by definition one of them has a resource
* fork tag on it. We need to compare the normalized filename before
* the ProDOS normalizer/uniqifier gets a chance to mangle it. As luck
* would have it, that's exactly what we have in "storageName".
*
* For a NuFX archive, NufxLib does all this nonsense for us, but we have
* to manage it ourselves here. The good news is that, since we have to
* wade through all the filenames, we have an opportunity to make the names
* unique. So long as we ensure that the names we have don't clash with
* anything currently on the disk, we know that anything we add that does
* clash is running into something we just added, which means we can turn
* on CreateFile's "make unique" feature and let the filesystem-specific
* code handle uniqueness.
*
Replace place-holder strings The NufxLib and diskimg libraries want narrow strings for certain things, notably for the "storage name", i.e. how the name will appear on the disk image or in the file archive. We need to convert from Windows UTF-16 to an Apple II filesystem-specific 8-bit character representation. We used to just pass narrow strings all the way through, so we didn't need any intermediate storage to hold the conversion. Now we do. In some cases there's nowhere good to put it. The initial UTF-16 conversion changes just dropped in some place-holder strings. This corrects the behavior, though in a couple of cases we're adding kluges on top of code that was already badly bent from its original intent (as initially conceived, CiderPress wasn't going to handle disk images, just ShrinkIt archives). It's not pretty, but it should work for now.
2014-12-05 01:05:48 +00:00
* Any fields we want to keep from the FileDetails struct need to be
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
* copied out. It's a "hairy" struct, so we need to duplicate the strings.
*/
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
NuError nuerr = kNuErrNone;
DiskFS* pDiskFS = pAddOpts->fpTargetDiskFS;
DIError dierr;
int neededLen = 64; // reasonable guess
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
char* fsNormalBuf = NULL; // name as it will appear on disk image
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGI(" +++ ADD file: orig='%ls' strip='%ls'",
(LPCWSTR) pDetails->GetLocalPathName(),
(LPCWSTR) pDetails->GetStrippedLocalPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
retry:
/*
* Convert "storageName" to a filesystem-normalized path.
*/
delete[] fsNormalBuf;
fsNormalBuf = new char[neededLen];
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
dierr = pDiskFS->NormalizePath(pDetails->GetStoragePathNameMOR(),
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
PathProposal::kDefaultStoredFssep, fsNormalBuf, &neededLen);
if (dierr == kDIErrDataOverrun) {
/* not long enough, try again *once* */
delete[] fsNormalBuf;
fsNormalBuf = new char[neededLen];
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
dierr = pDiskFS->NormalizePath(pDetails->GetStoragePathNameMOR(),
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
PathProposal::kDefaultStoredFssep, fsNormalBuf, &neededLen);
}
if (dierr != kDIErrNone) {
nuerr = kNuErrInternal;
goto bail;
}
/*
* Test to see if the file already exists. If it does, give the user
* the opportunity to rename it, overwrite the original, or skip
* adding it.
*
* The FS-normalized path may not reflect the actual storage name,
* because some features (like ProDOS "allow lower case") aren't
* factored in until later. However, it should be close enough -- it
* has to be, or we'd be in trouble for saying it's going to overwrite
* the file in the archive.
*/
A2File* pExisting;
pExisting = pDiskFS->GetFileByName(fsNormalBuf);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pExisting != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
NuResult result;
result = HandleReplaceExisting(pExisting, pDetails);
if (result == kNuAbort) {
nuerr = kNuErrAborted;
goto bail;
} else if (result == kNuSkip) {
nuerr = kNuErrSkipped;
goto bail;
} else if (result == kNuRename) {
goto retry;
} else if (result == kNuOverwrite) {
/* delete the existing file immediately */
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD(" Deleting existing file '%hs'", fsNormalBuf);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dierr = pDiskFS->DeleteFile(pExisting);
if (dierr != kDIErrNone) {
// Would be nice to show a dialog and explain *why*, but
// I'm not sure we have a window here.
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGE(" Deletion failed (err=%d)", dierr);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
} else {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGE("GLITCH: bad return %d from HandleReplaceExisting",result);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
assert(false);
nuerr = kNuErrInternal;
goto bail;
}
}
/*
* Put all the goodies into a new FileAddData object, and add it to
* the end of the list.
*/
FileAddData* pAddData;
pAddData = new FileAddData(pDetails, fsNormalBuf);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pAddData == NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
nuerr = kNuErrMalloc;
goto bail;
}
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD("FSNormalized is '%hs'", pAddData->GetFSNormalPath());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
AddToAddDataList(pAddData);
bail:
delete[] fsNormalBuf;
return nuerr;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
NuResult DiskArchive::HandleReplaceExisting(const A2File* pExisting,
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LocalFileDetails* pDetails)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
NuResult result;
if (fOverwriteNoAsk) {
if (fOverwriteExisting)
return kNuOverwrite;
else
return kNuSkip;
}
ConfirmOverwriteDialog confOvwr;
Update progress dialogs Do the Mac OS Roman conversion in a couple of places. Fix the NuFX extraction progress, which got all blinky after the wide char update.
2015-01-15 23:33:12 +00:00
confOvwr.fExistingFile = Charset::ConvertMORToUNI(pExisting->GetPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
confOvwr.fExistingFileModWhen = pExisting->GetModWhen();
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
PathName srcPath(pDetails->GetLocalPathName());
confOvwr.fNewFileSource = pDetails->GetLocalPathName();
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
confOvwr.fNewFileModWhen = srcPath.GetModWhen();
if (confOvwr.DoModal() == IDCANCEL) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("User cancelled out of add-to-diskimg replace-existing");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
return kNuAbort;
}
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
// TODO: if they rename one fork, we need to track that fact and
// carry it over to the other fork -- otherwise they can rename
// the data and resource forks into separate files.
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (confOvwr.fResultRename) {
/*
* Replace the name in FileDetails. They were asked to modify
* the already-normalized version of the filename. We will run
* it back through the FS-specific normalizer, which will handle
* any oddities they type in.
*
* We don't want to run it through PathProposal.LocalToArchive
* because that'll strip out ':' in the pathnames.
*
* Ideally the rename dialog would have a way to validate the
* full path and reject "OK" if it's not valid. Instead, we just
* allow the FS normalizer to force the filename to be valid.
*/
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
pDetails->SetStrippedLocalPathName(confOvwr.fExistingFile);
LOGI("Trying rename to '%ls'",
(LPCWSTR) pDetails->GetStrippedLocalPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
return kNuRename;
}
if (confOvwr.fResultApplyToAll) {
fOverwriteNoAsk = true;
if (confOvwr.fResultOverwrite)
fOverwriteExisting = true;
else
fOverwriteExisting = false;
}
if (confOvwr.fResultOverwrite)
result = kNuOverwrite;
else
result = kNuSkip;
return result;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
CString DiskArchive::ProcessFileAddData(DiskFS* pDiskFS, int addOptsConvEOL)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
CString errMsg;
FileAddData* pData;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
unsigned char* dataBuf = NULL;
unsigned char* rsrcBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
long dataLen, rsrcLen;
MainWindow* pMainWin = (MainWindow*)::AfxGetMainWnd();
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("--- ProcessFileAddData");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/* map the EOL conversion to something we can use */
GenericEntry::ConvertEOL convEOL;
switch (addOptsConvEOL) {
case AddFilesDialog::kConvEOLNone:
convEOL = GenericEntry::kConvertEOLOff;
break;
case AddFilesDialog::kConvEOLType:
// will be adjusted each time through the loop
convEOL = GenericEntry::kConvertEOLOff;
break;
case AddFilesDialog::kConvEOLAuto:
convEOL = GenericEntry::kConvertEOLAuto;
break;
case AddFilesDialog::kConvEOLAll:
convEOL = GenericEntry::kConvertEOLOn;
break;
default:
assert(false);
convEOL = GenericEntry::kConvertEOLOff;
break;
}
pData = fpAddDataHead;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
while (pData != NULL) {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
const LocalFileDetails* pDataDetails = NULL;
const LocalFileDetails* pRsrcDetails = NULL;
const LocalFileDetails* pDetails = pData->GetDetails();
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
const char* typeStr = "????"; // for debug msg only
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
switch (pDetails->GetEntryKind()) {
case LocalFileDetails::kFileKindDataFork:
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pDataDetails = pDetails;
typeStr = "data";
break;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
case LocalFileDetails::kFileKindRsrcFork:
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pRsrcDetails = pDetails;
typeStr = "rsrc";
break;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
case LocalFileDetails::kFileKindDiskImage:
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pDataDetails = pDetails;
typeStr = "disk";
break;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
case LocalFileDetails::kFileKindBothForks:
case LocalFileDetails::kFileKindDirectory:
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
default:
assert(false);
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
return L"internal error";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pData->GetOtherFork() != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pDetails = pData->GetOtherFork()->GetDetails();
typeStr = "both";
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
switch (pDetails->GetEntryKind()) {
case LocalFileDetails::kFileKindDataFork:
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
assert(pDataDetails == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pDataDetails = pDetails;
break;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
case LocalFileDetails::kFileKindRsrcFork:
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
assert(pRsrcDetails == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pRsrcDetails = pDetails;
break;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
case LocalFileDetails::kFileKindDiskImage:
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
assert(false);
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
return L"(internal) add other disk error";
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
case LocalFileDetails::kFileKindBothForks:
case LocalFileDetails::kFileKindDirectory:
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
default:
assert(false);
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
return L"internal error";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
}
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Adding file '%ls' (%hs)",
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
(LPCWSTR) pDetails->GetStrippedLocalPathName(), typeStr);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pDataDetails != NULL || pRsrcDetails != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* The current implementation of DiskImg/DiskFS requires writing each
* fork in one shot. This means loading the entire thing into
* memory. Not so bad for ProDOS, with its 16MB maximum file size,
* but it could be awkward for HFS (not to mention HFS Plus!).
*/
DiskFS::CreateParms parms;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
/* use the FS-normalized path here */
/* (do we have to? do we want to?) */
parms.pathName = pData->GetFSNormalPath();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pRsrcDetails != NULL)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
parms.storageType = kNuStorageExtended;
else
parms.storageType = kNuStorageSeedling;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
/* copy the rest out of the LocalFileDetails */
parms.fssep = pDetails->GetFssep();
parms.fileType = pDetails->GetFileType();
parms.auxType = pDetails->GetExtraType();
parms.access = pDetails->GetAccess();
parms.createWhen = NufxArchive::DateTimeToSeconds(&pDetails->GetCreateWhen());
parms.modWhen = NufxArchive::DateTimeToSeconds(&pDetails->GetModWhen());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dataLen = rsrcLen = -1;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pDataDetails != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/* figure out text conversion, including high ASCII for DOS */
/* (HA conversion only happens if text conversion happens) */
GenericEntry::ConvertHighASCII convHA;
if (addOptsConvEOL == AddFilesDialog::kConvEOLType) {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
if (pDataDetails->GetFileType() == kFileTypeTXT ||
pDataDetails->GetFileType() == kFileTypeSRC)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Enabling text conversion by type");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
convEOL = GenericEntry::kConvertEOLOn;
} else {
convEOL = GenericEntry::kConvertEOLOff;
}
}
if (DiskImg::UsesDOSFileStructure(pDiskFS->GetDiskImg()->GetFSFormat()))
convHA = GenericEntry::kConvertHAOn;
else
convHA = GenericEntry::kConvertHAOff;
Allow files larger than 16MB to be added to HFS volumes Ideally this would add a "what's your limit" call to the disk image library, but this will do for now.
2020-03-26 18:08:24 +00:00
errMsg = LoadFile(pDataDetails->GetLocalPathName(), pDiskFS,
&dataBuf, &dataLen, convEOL, convHA);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (!errMsg.IsEmpty())
goto bail;
}
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pRsrcDetails != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/* no text conversion on resource forks */
Allow files larger than 16MB to be added to HFS volumes Ideally this would add a "what's your limit" call to the disk image library, but this will do for now.
2020-03-26 18:08:24 +00:00
errMsg = LoadFile(pRsrcDetails->GetLocalPathName(), pDiskFS,
&rsrcBuf, &rsrcLen,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
GenericEntry::kConvertEOLOff, GenericEntry::kConvertHAOff);
if (!errMsg.IsEmpty())
goto bail;
}
/* really ought to do this separately for each thread */
SET_PROGRESS_BEGIN();
Update progress dialogs Do the Mac OS Roman conversion in a couple of places. Fix the NuFX extraction progress, which got all blinky after the wide char update.
2015-01-15 23:33:12 +00:00
CString pathNameW(Charset::ConvertMORToUNI(parms.pathName));
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
SET_PROGRESS_UPDATE2(0, pDetails->GetLocalPathName(), pathNameW);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DIError dierr;
dierr = AddForksToDisk(pDiskFS, &parms, dataBuf, dataLen,
rsrcBuf, rsrcLen);
SET_PROGRESS_END();
if (dierr != kDIErrNone) {
Update progress dialogs Do the Mac OS Roman conversion in a couple of places. Fix the NuFX extraction progress, which got all blinky after the wide char update.
2015-01-15 23:33:12 +00:00
errMsg.Format(L"Unable to add '%ls' to image: %hs.",
(LPCWSTR) pathNameW, DiskImgLib::DIStrError(dierr));
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
delete[] dataBuf;
delete[] rsrcBuf;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
dataBuf = rsrcBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pData = pData->GetNext();
}
bail:
delete[] dataBuf;
delete[] rsrcBuf;
return errMsg;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
// TODO: really ought to update the progress counter, especially when reading
// really large files.
Allow files larger than 16MB to be added to HFS volumes Ideally this would add a "what's your limit" call to the disk image library, but this will do for now.
2020-03-26 18:08:24 +00:00
// TODO: shouldn't be loading really large files into memory. ProDOS is fine
// (16MB limit), HFS could be problematic in a 32-bit app.
CString DiskArchive::LoadFile(const WCHAR* pathName, DiskFS* pDiskFS,
uint8_t** pBuf, long* pLen,
GenericEntry::ConvertEOL conv, GenericEntry::ConvertHighASCII convHA) const
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
const char kCharLF = '\n';
const char kCharCR = '\r';
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
CString errMsg;
FILE* fp;
long fileLen;
ASSERT(convHA == GenericEntry::kConvertHAOn ||
convHA == GenericEntry::kConvertHAOff);
ASSERT(conv == GenericEntry::kConvertEOLOn ||
conv == GenericEntry::kConvertEOLOff ||
conv == GenericEntry::kConvertEOLAuto);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pathName != NULL);
ASSERT(pBuf != NULL);
ASSERT(pLen != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
fp = _wfopen(pathName, L"rb");
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (fp == NULL) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to open '%ls': %hs.", pathName,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
strerror(errno));
goto bail;
}
if (fseek(fp, 0, SEEK_END) != 0) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to seek to end of '%ls': %hs", pathName,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
strerror(errno));
goto bail;
}
fileLen = ftell(fp);
if (fileLen < 0) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to determine length of '%ls': %hs", pathName,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
strerror(errno));
goto bail;
}
rewind(fp);
if (fileLen == 0) { // handle zero-length files
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
*pBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
*pLen = 0;
goto bail;
Allow files larger than 16MB to be added to HFS volumes Ideally this would add a "what's your limit" call to the disk image library, but this will do for now.
2020-03-26 18:08:24 +00:00
}
// TODO: we want to limit the file length based on the filesystem type,
// so we don't initiate an operation that can't possibly succeed. We
// should be able to query the DiskImg for its max file size, but that
// API doesn't currently exist. For now, limit anything other than HFS
// to 16MB.
if (pDiskFS->GetDiskImg()->GetFSFormat() != DiskImg::kFormatMacHFS &&
fileLen > 0x00ffffff) {
errMsg = L"Cannot add files larger than 16MB to a non-HFS disk image.";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
Use types with explicit sizes Much of what the "reformat" code does involves processing data that is 8, 16, or 32 bits. We want to use size-specific types from stdint.h (e.g. uint16_t) rather than "unsigned short". This was a quick pass to replace the various "unsigned" declarations. More can be done here and elsewhere.
2014-11-21 02:10:18 +00:00
*pBuf = new uint8_t[fileLen];
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (*pBuf == NULL) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to allocate %ld bytes for '%ls'.",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
fileLen, pathName);
goto bail;
}
/*
* We're ready to load the file. We need to sort out EOL conversion.
* Since we always convert to CR, we know the file will stay the same
* size or get smaller, which means the buffer we've allocated is
* guaranteed to hold the file even if we convert it.
*
* If the text mode is "auto", we need to load a piece of the file and
* analyze it.
*/
if (conv == GenericEntry::kConvertEOLAuto) {
GenericEntry::EOLType eolType;
GenericEntry::ConvertHighASCII dummy;
int chunkLen = 16384; // nice big chunk
if (chunkLen > fileLen)
chunkLen = fileLen;
if (fread(*pBuf, chunkLen, 1, fp) != 1) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to read initial chunk of '%ls': %hs.",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pathName, strerror(errno));
delete[] *pBuf;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
*pBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
rewind(fp);
conv = GenericEntry::DetermineConversion(*pBuf, chunkLen,
&eolType, &dummy);
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("LoadFile DetermineConv returned conv=%d eolType=%d",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
conv, eolType);
if (conv == GenericEntry::kConvertEOLOn &&
eolType == GenericEntry::kEOLCR)
{
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" (skipping conversion due to matching eolType)");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
conv = GenericEntry::kConvertEOLOff;
}
}
ASSERT(conv != GenericEntry::kConvertEOLAuto);
/*
* The "high ASCII" conversion is either on or off. In this context,
* "on" means "convert all text files", and "off" means "don't convert
* text files". We never convert non-text files. Conversion should
* always be "on" for DOS 3.2/3.3, and "off" for everything else (except
* RDOS, should we choose to make that writeable).
*/
if (conv == GenericEntry::kConvertEOLOff) {
/* fast path */
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" +++ NOT converting text '%ls'", pathName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (fread(*pBuf, fileLen, 1, fp) != 1) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to read '%ls': %hs.", pathName, strerror(errno));
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
delete[] *pBuf;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
*pBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
} else {
/*
* Convert as we go.
*
* Observation: if we copy a binary file to a DOS disk, and force
* the text conversion, we will convert 0x0a to 0x0d, and thence
* to 0x8d. However, we may still have some 0x8a bytes lying around,
* because we don't convert 0x8a in the original file to anything.
* This means that a CR->CRLF or LF->CRLF conversion can't be
* "undone" on a DOS disk. (Not that anyone cares.)
*/
long count = fileLen;
int mask, ic;
bool lastCR = false;
unsigned char* buf = *pBuf;
if (convHA == GenericEntry::kConvertHAOn)
mask = 0x80;
else
mask = 0x00;
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" +++ Converting text '%ls', mask=0x%02x", pathName, mask);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
while (count--) {
ic = getc(fp);
if (ic == kCharCR) {
*buf++ = (unsigned char) (kCharCR | mask);
lastCR = true;
} else if (ic == kCharLF) {
if (!lastCR)
*buf++ = (unsigned char) (kCharCR | mask);
lastCR = false;
} else {
if (ic == '\0')
*buf++ = (unsigned char) ic; // don't conv 0x00
else
*buf++ = (unsigned char) (ic | mask);
lastCR = false;
}
}
fileLen = buf - *pBuf;
}
(void) fclose(fp);
*pLen = fileLen;
bail:
return errMsg;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
DIError DiskArchive::AddForksToDisk(DiskFS* pDiskFS,
const DiskFS::CreateParms* pParms,
Fix a few compiler warnings Mostly uninitialized class members. Should not cause a change in behavior.
2021-05-10 01:17:49 +00:00
const uint8_t* dataBuf, long dataLen,
const uint8_t* rsrcBuf, long rsrcLen) const
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
DIError dierr = kDIErrNone;
const int kFileTypeBIN = 0x06;
const int kFileTypeINT = 0xfa;
const int kFileTypeBAS = 0xfc;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
A2File* pNewFile = NULL;
A2FileDescr* pOpenFile = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskFS::CreateParms parmCopy;
/*
* Make a temporary copy, pointers and all, so we can rewrite some of
* the fields. This is sort of bad, because we're making copies of a
* const char* filename pointer whose underlying storage we're not
* really familiar with. However, so long as we don't try to retain
* it after this function returns we should be fine.
*
* Might be better to make CreateParms a class instead of a struct,
* make the pathName field new[] storage, and write a copy constructor
* for the operation below. This will be fine for now.
*/
memcpy(&parmCopy, pParms, sizeof(parmCopy));
if (rsrcLen >= 0) {
ASSERT(parmCopy.storageType == kNuStorageExtended);
}
/*
* Look for "empty directory holders" that we put into NuFX archives
* when doing disk-to-archive conversions. These make no sense if
* there's no fssep (because it's coming from DOS), or if there's no
* base path, so we can ignore those cases. We can also ignore it if
* the file is forked or is already a directory.
*/
if (parmCopy.fssep != '\0' && parmCopy.storageType == kNuStorageSeedling) {
const char* cp;
cp = strrchr(parmCopy.pathName, parmCopy.fssep);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (cp != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (strcmp(cp+1, kEmptyFolderMarker) == 0 && dataLen == 0) {
/* drop the junk on the end */
parmCopy.storageType = kNuStorageDirectory;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
CStringA replacementFileName(parmCopy.pathName);;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
replacementFileName =
replacementFileName.Left(cp - parmCopy.pathName -1);
parmCopy.pathName = replacementFileName;
parmCopy.fileType = 0x0f; // DIR
parmCopy.access &= ~(A2FileProDOS::kAccessInvisible);
dataLen = -1;
}
}
}
/*
* If this is a subdir create request (from the clipboard or an "empty
* directory placeholder" in a NuFX archive), handle it here. If we're
* on a filesystem that doesn't have subdirectories, just skip it.
*/
if (parmCopy.storageType == kNuStorageDirectory) {
A2File* pDummyFile;
ASSERT(dataLen < 0 && rsrcLen < 0);
if (DiskImg::IsHierarchical(pDiskFS->GetDiskImg()->GetFSFormat())) {
dierr = pDiskFS->CreateFile(&parmCopy, &pDummyFile);
if (dierr == kDIErrDirectoryExists)
dierr = kDIErrNone; // dirs are not made unique
goto bail;
} else {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" Ignoring subdir create req on non-hierarchic FS");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
}
/* don't try to put resource forks onto a DOS disk */
if (!DiskImg::HasResourceForks(pDiskFS->GetDiskImg()->GetFSFormat())) {
if (rsrcLen >= 0) {
rsrcLen = -1;
parmCopy.storageType = kNuStorageSeedling;
if (dataLen < 0) {
/* this was a resource-fork-only file */
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("--- nothing left to write for '%hs'",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
parmCopy.pathName);
goto bail;
}
} else {
ASSERT(parmCopy.storageType == kNuStorageSeedling);
}
}
/* quick kluge to get the right file type on large DOS files */
if (DiskImg::UsesDOSFileStructure(pDiskFS->GetDiskImg()->GetFSFormat()) &&
dataLen >= 65536)
{
if (parmCopy.fileType == kFileTypeBIN ||
parmCopy.fileType == kFileTypeINT ||
parmCopy.fileType == kFileTypeBAS)
{
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("+++ switching DOS file type to $f2");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
parmCopy.fileType = 0xf2; // DOS 'S' file
}
}
/*
* Create the file on the disk. The storage type determines whether
* it has data+rsrc forks or just data (there's no such thing in
* ProDOS as "just a resource fork"). There's no need to open the
* fork if we're not going to write to it.
*
* This holds for resource forks as well, because the storage type
* determines whether or not the file is forked, and we've asserted
* that a file with a non-(-1) rsrcLen is forked.
*/
dierr = pDiskFS->CreateFile(&parmCopy, &pNewFile);
if (dierr != kDIErrNone) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" CreateFile failed: %hs", DiskImgLib::DIStrError(dierr));
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
/*
* Note: if this was an empty directory holder, pNewFile will be set
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
* to NULL. We used to avoid handling this by just not opening the file
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
* if it had a length of zero. However, DOS 3.3 needs to write some
* kinds of zero-length files, because e.g. a zero-length 'B' file
* actually has 4 bytes of data in it.
*
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
* Of course, if dataLen is zero then dataBuf is NULL, so we need to
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
* supply a dummy write buffer. None of this is an issue for resource
* forks, because DOS 3.3 doesn't have those.
*/
if (dataLen > 0 ||
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
(dataLen == 0 && pNewFile != NULL))
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pNewFile != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
unsigned char dummyBuf[1] = { '\0' };
dierr = pNewFile->Open(&pOpenFile, false, false);
if (dierr != kDIErrNone)
goto bail;
pOpenFile->SetProgressUpdater(DiskArchive::ProgressCallback,
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
dataLen, NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
dierr = pOpenFile->Write(dataBuf != NULL ? dataBuf : dummyBuf, dataLen);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (dierr != kDIErrNone)
goto bail;
dierr = pOpenFile->Close();
if (dierr != kDIErrNone)
goto bail;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
pOpenFile = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
if (rsrcLen > 0) {
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pNewFile != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dierr = pNewFile->Open(&pOpenFile, false, true);
if (dierr != kDIErrNone)
goto bail;
pOpenFile->SetProgressUpdater(DiskArchive::ProgressCallback,
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
rsrcLen, NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dierr = pOpenFile->Write(rsrcBuf, rsrcLen);
if (dierr != kDIErrNone)
goto bail;
dierr = pOpenFile->Close();
if (dierr != kDIErrNone)
goto bail;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
pOpenFile = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
bail:
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (pOpenFile != NULL)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pOpenFile->Close();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (dierr != kDIErrNone && pNewFile != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* Clean up the partially-written file. This does not, of course,
* erase any subdirectories that were created to contain this file.
* Not worth worrying about.
*/
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" Deleting newly-created file '%hs'", parmCopy.pathName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
(void) pDiskFS->DeleteFile(pNewFile);
}
return dierr;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
void DiskArchive::AddToAddDataList(FileAddData* pData)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pData != NULL);
ASSERT(pData->GetNext() == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* Run through the entire existing list, looking for a match. This is
* O(n^2) behavior, but I'm expecting N to be relatively small (under
* 1000 in almost all cases).
*/
FileAddData* pSearch = fpAddDataHead;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LocalFileDetails::FileKind dataKind, listKind;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
dataKind = pData->GetDetails()->GetEntryKind();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
while (pSearch != NULL) {
if (pSearch->GetOtherFork() == NULL &&
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
wcscmp(pSearch->GetDetails()->GetStrippedLocalPathName(),
pData->GetDetails()->GetStrippedLocalPathName()) == 0)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
//NuThreadID dataID = pData->GetDetails()->threadID;
//NuThreadID listID = pSearch->GetDetails()->threadID;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
listKind = pSearch->GetDetails()->GetEntryKind();
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/* got a name match */
if (dataKind != listKind &&
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
(dataKind == LocalFileDetails::kFileKindDataFork ||
dataKind == LocalFileDetails::kFileKindRsrcFork) &&
(listKind == LocalFileDetails::kFileKindDataFork ||
listKind == LocalFileDetails::kFileKindRsrcFork))
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
/* looks good, hook it in here instead of the list */
Fix some issues identified by static analyzer Includes "printf format" annotation on debug log function.
2014-11-19 22:54:24 +00:00
LOGD("--- connecting forks of '%ls' and '%ls'",
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
(LPCWSTR) pData->GetDetails()->GetLocalPathName(),
(LPCWSTR) pSearch->GetDetails()->GetLocalPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pSearch->SetOtherFork(pData);
return;
}
}
pSearch = pSearch->GetNext();
}
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (fpAddDataHead == NULL) {
assert(fpAddDataTail == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
fpAddDataHead = fpAddDataTail = pData;
} else {
fpAddDataTail->SetNext(pData);
fpAddDataTail = pData;
}
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
void DiskArchive::FreeAddDataList(void)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
FileAddData* pData;
FileAddData* pNext;
pData = fpAddDataHead;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
while (pData != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pNext = pData->GetNext();
delete pData->GetOtherFork();
delete pData;
pData = pNext;
}
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
fpAddDataHead = fpAddDataTail = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
/*
* ===========================================================================
* DiskArchive -- create subdir
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskArchive::CreateSubdir(CWnd* pMsgWnd, GenericEntry* pParentEntry,
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
const WCHAR* newName)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(newName != NULL && wcslen(newName) > 0);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskEntry* pEntry = (DiskEntry*) pParentEntry;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pEntry != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
A2File* pFile = pEntry->GetA2File();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pFile != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskFS* pDiskFS = pFile->GetDiskFS();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pDiskFS != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (!pFile->IsDirectory()) {
ASSERT(false);
return false;
}
DIError dierr;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
A2File* pNewFile = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskFS::CreateParms parms;
Update progress dialogs Do the Mac OS Roman conversion in a couple of places. Fix the NuFX extraction progress, which got all blinky after the wide char update.
2015-01-15 23:33:12 +00:00
CStringA pathNameMOR;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
time_t now = time(NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* Create the full path.
*/
if (pFile->IsVolumeDirectory()) {
Update progress dialogs Do the Mac OS Roman conversion in a couple of places. Fix the NuFX extraction progress, which got all blinky after the wide char update.
2015-01-15 23:33:12 +00:00
pathNameMOR = newName;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
} else {
Update progress dialogs Do the Mac OS Roman conversion in a couple of places. Fix the NuFX extraction progress, which got all blinky after the wide char update.
2015-01-15 23:33:12 +00:00
pathNameMOR = pParentEntry->GetPathNameMOR();
pathNameMOR += pParentEntry->GetFssep();
pathNameMOR += newName;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(wcschr(newName, pParentEntry->GetFssep()) == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/* using NufxLib constants; they match with ProDOS */
memset(&parms, 0, sizeof(parms));
Update progress dialogs Do the Mac OS Roman conversion in a couple of places. Fix the NuFX extraction progress, which got all blinky after the wide char update.
2015-01-15 23:33:12 +00:00
parms.pathName = pathNameMOR;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
parms.fssep = pParentEntry->GetFssep();
parms.storageType = kNuStorageDirectory;
parms.fileType = 0x0f; // ProDOS DIR
parms.auxType = 0;
parms.access = kNuAccessUnlocked;
parms.createWhen = now;
parms.modWhen = now;
dierr = pDiskFS->CreateFile(&parms, &pNewFile);
if (dierr != kDIErrNone) {
CString errMsg;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to create subdirectory: %hs.\n",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
return false;
}
if (InternalReload(pMsgWnd) != 0)
return false;
return true;
}
/*
* ===========================================================================
* DiskArchive -- delete selection
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*static*/ int DiskArchive::CompareDisplayNamesDesc(const void* ventry1,
const void* ventry2)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
const DiskEntry* pEntry1 = *((const DiskEntry**) ventry1);
const DiskEntry* pEntry2 = *((const DiskEntry**) ventry2);
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
return wcsicmp(pEntry2->GetDisplayName(), pEntry1->GetDisplayName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskArchive::DeleteSelection(CWnd* pMsgWnd, SelectionSet* pSelSet)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*
* The DiskFS DeleteFile() function will not delete a subdirectory unless
* it is empty. This complicates matters somewhat for us, because the
* selection set isn't in any particular order. We need to sort on the
* pathname and then delete bottom-up.
*
* CiderPress does work to ensure that, if a subdir is selected, everything
* in that subdir is also selected. So if we just delete everything in the
* right order, we should be okay.
*/
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
CString errMsg;
SelectionEntry* pSelEntry;
DiskEntry* pEntry;
DIError dierr;
bool retVal = false;
SET_PROGRESS_BEGIN();
/*
* Start by copying the DiskEntry pointers out of the selection set and
* into an array. The selection set was created such that there is one
* entry in the set for each file. (The file viewer likes to have one
* entry for each thread.)
*/
int numEntries = pSelSet->GetNumEntries();
ASSERT(numEntries > 0);
DiskEntry** entryArray = new DiskEntry*[numEntries];
int idx = 0;
pSelEntry = pSelSet->IterNext();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
while (pSelEntry != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pEntry = (DiskEntry*) pSelEntry->GetEntry();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pEntry != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
entryArray[idx++] = pEntry;
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Added 0x%08lx '%ls'", (long) entryArray[idx-1],
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
(LPCWSTR) entryArray[idx-1]->GetDisplayName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pSelEntry = pSelSet->IterNext();
}
ASSERT(idx == numEntries);
/*
* Sort the file array by descending filename.
*/
::qsort(entryArray, numEntries, sizeof(DiskEntry*), CompareDisplayNamesDesc);
/*
* Run through the sorted list, deleting each entry.
*/
for (idx = 0; idx < numEntries; idx++) {
A2File* pFile;
pEntry = entryArray[idx];
pFile = pEntry->GetA2File();
/*
* We shouldn't be here at all if the main volume were opened
* read-only. However, it's possible that the main is read-write
* and our sub-volumes are read-only (probably because we don't
* support write access to the filesystem).
*/
if (!pFile->GetDiskFS()->GetReadWriteSupported()) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to delete '%ls' on '%hs': operation not supported.",
(LPCWSTR) pEntry->GetDisplayName(),
(LPCSTR) pFile->GetDiskFS()->GetVolumeName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
goto bail;
}
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
LOGI(" Deleting '%ls' from '%hs'", (LPCWSTR) pEntry->GetPathNameUNI(),
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
(LPCSTR) pFile->GetDiskFS()->GetVolumeName());
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
// TODO: should be using display name for progress updater?
SET_PROGRESS_UPDATE2(0, pEntry->GetPathNameUNI(), NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* Ask the DiskFS to delete the file. As soon as this completes,
* "pFile" is invalid and must not be dereferenced.
*/
dierr = pFile->GetDiskFS()->DeleteFile(pFile);
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to delete '%ls' on '%hs': %hs.",
(LPCWSTR) pEntry->GetDisplayName(),
(LPCSTR) pFile->GetDiskFS()->GetVolumeName(),
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
goto bail;
}
SET_PROGRESS_UPDATE(100);
/*
* Be paranoid and zap the pointer, on the off chance somebody
* tries to redraw the content list from the deleted data.
*
* In practice we don't work this way -- the stuff that gets drawn
* on the screen comes out of GenericEntry, not A2File. If this
* changes we'll need to raise the "reload" flag here, before the
* reload, to prevent the ContentList from chasing a bad pointer.
*/
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
pEntry->SetA2File(NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
retVal = true;
bail:
SET_PROGRESS_END();
delete[] entryArray;
if (InternalReload(pMsgWnd) != 0)
retVal = false;
return retVal;
}
/*
* ===========================================================================
* DiskArchive -- rename files
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskArchive::RenameSelection(CWnd* pMsgWnd, SelectionSet* pSelSet)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*
* If we rename a subdirectory, it could affect the next thing we try to
* rename (because we show the full path). We have to reload our file
* list from the DiskFS after each renamed subdir. The trouble is that
* this invalidates the data displayed in the ContentList, and we won't
* redraw the screen correctly. We can work around the problem by getting
* the pathname directly from the DiskFS instead of from DiskEntry, though
* it's not immediately obvious which is less confusing.
*/
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
CString errMsg;
bool retVal = false;
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Renaming %d entries", pSelSet->GetNumEntries());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* For each item in the selection set, bring up the "rename" dialog,
* and ask the GenericEntry to process it.
*
* If they hit "cancel" or there's an error, we still flush the
* previous changes. This is so that we don't have to create the
* same sort of deferred-write feature when renaming things in other
* sorts of archives (e.g. disk archives).
*/
SelectionEntry* pSelEntry = pSelSet->IterNext();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
while (pSelEntry != NULL) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
RenameEntryDialog renameDlg(pMsgWnd);
DiskEntry* pEntry = (DiskEntry*) pSelEntry->GetEntry();
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
LOGI(" Renaming '%ls'", (LPCWSTR) pEntry->GetPathNameUNI());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (!SetRenameFields(pMsgWnd, pEntry, &renameDlg))
break;
int result;
if (pEntry->GetA2File()->IsVolumeDirectory())
result = IDIGNORE; // don't allow rename of volume dir
else
result = renameDlg.DoModal();
if (result == IDOK) {
DIError dierr;
DiskFS* pDiskFS;
A2File* pFile;
pFile = pEntry->GetA2File();
pDiskFS = pFile->GetDiskFS();
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
CStringA newNameA(renameDlg.fNewName);
dierr = pDiskFS->RenameFile(pFile, newNameA);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to rename '%ls' to '%ls': %hs.",
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
(LPCWSTR) pEntry->GetPathNameUNI(),
(LPCWSTR) renameDlg.fNewName,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
goto bail;
}
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD("Rename of '%ls' to '%ls' succeeded",
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
(LPCWSTR) pEntry->GetDisplayName(),
(LPCWSTR) renameDlg.fNewName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
} else if (result == IDCANCEL) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Canceling out of remaining renames");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
} else {
/* 3rd possibility is IDIGNORE, i.e. skip this entry */
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
LOGI("Skipping rename of '%ls'", (LPCWSTR) pEntry->GetDisplayName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
pSelEntry = pSelSet->IterNext();
}
/* reload GenericArchive from disk image */
if (InternalReload(pMsgWnd) == kNuErrNone)
retVal = true;
bail:
return retVal;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskArchive::SetRenameFields(CWnd* pMsgWnd, DiskEntry* pEntry,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
RenameEntryDialog* pDialog)
{
DiskFS* pDiskFS;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pEntry != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
/*
* Figure out if we're allowed to change the entire path. (This is
* doing it the hard way, but what the hell.)
*/
long cap = GetCapability(GenericArchive::kCapCanRenameFullPath);
bool renameFullPath = (cap != 0);
// a bit round-about, but it works
pDiskFS = pEntry->GetA2File()->GetDiskFS();
/*
* Make sure rename is allowed. It's nice to do these *before* putting
* up the rename dialog, so that the user doesn't do a bunch of typing
* before being told that it's pointless.
*/
if (!pDiskFS->GetReadWriteSupported()) {
CString errMsg;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to rename '%ls': operation not supported.",
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
(LPCWSTR) pEntry->GetPathNameUNI());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
return false;
}
if (pDiskFS->GetFSDamaged()) {
CString errMsg;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to rename '%ls': the disk it's on appears to be damaged.",
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
(LPCWSTR) pEntry->GetPathNameUNI());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
return false;
}
pDialog->SetCanRenameFullPath(renameFullPath);
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
pDialog->fOldName = pEntry->GetPathNameUNI();
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pDialog->fFssep = pEntry->GetFssep();
pDialog->fpArchive = this;
pDialog->fpEntry = pEntry;
return true;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
CString DiskArchive::TestPathName(const GenericEntry* pGenericEntry,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
const CString& basePath, const CString& newName, char newFssep) const
{
const DiskEntry* pEntry = (DiskEntry*) pGenericEntry;
DiskImg::FSFormat format;
Minor fixes Fix release build. Eliminate a few warnings.
2014-11-18 22:30:51 +00:00
CString errMsg, pathName;
CStringA newNameA;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskFS* pDiskFS;
if (basePath.IsEmpty()) {
pathName = newName;
} else {
pathName = basePath;
pathName += newFssep;
pathName += newName;
}
pDiskFS = pEntry->GetA2File()->GetDiskFS();
format = pDiskFS->GetDiskImg()->GetFSFormat();
/* look for an existing file, but don't compare against self */
A2File* existingFile;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
CStringA pathNameA(pathName);
existingFile = pDiskFS->GetFileByName(pathNameA);
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (existingFile != NULL && existingFile != pEntry->GetA2File()) {
Minor fixes Fix release build. Eliminate a few warnings.
2014-11-18 22:30:51 +00:00
errMsg = L"A file with that name already exists.";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
newNameA = newName;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
switch (format) {
case DiskImg::kFormatProDOS:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (!DiskFSProDOS::IsValidFileName(newNameA))
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&errMsg, IDS_VALID_FILENAME_PRODOS);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatDOS33:
case DiskImg::kFormatDOS32:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (!DiskFSDOS33::IsValidFileName(newNameA))
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&errMsg, IDS_VALID_FILENAME_DOS);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatPascal:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (!DiskFSPascal::IsValidFileName(newNameA))
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&errMsg, IDS_VALID_FILENAME_PASCAL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatMacHFS:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (!DiskFSHFS::IsValidFileName(newNameA))
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&errMsg, IDS_VALID_FILENAME_HFS);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
default:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg = L"Not supported by TestPathName!";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
bail:
return errMsg;
}
/*
* ===========================================================================
* DiskArchive -- rename a volume
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskArchive::RenameVolume(CWnd* pMsgWnd, DiskFS* pDiskFS,
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
const WCHAR* newName)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
DIError dierr;
CString errMsg;
bool retVal = true;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
CStringA newNameA(newName);
dierr = pDiskFS->RenameVolume(newNameA);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to rename volume: %hs.\n",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
retVal = false;
/* fall through to reload anyway */
}
/* reload GenericArchive from disk image */
if (InternalReload(pMsgWnd) != 0)
retVal = false;
return retVal;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
CString DiskArchive::TestVolumeName(const DiskFS* pDiskFS,
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
const WCHAR* newName) const
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
DiskImg::FSFormat format;
CString errMsg;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pDiskFS != NULL);
ASSERT(newName != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
format = pDiskFS->GetDiskImg()->GetFSFormat();
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
CStringA newNameA(newName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
switch (format) {
case DiskImg::kFormatProDOS:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (!DiskFSProDOS::IsValidVolumeName(newNameA))
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&errMsg, IDS_VALID_VOLNAME_PRODOS);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatDOS33:
case DiskImg::kFormatDOS32:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (!DiskFSDOS33::IsValidVolumeName(newNameA))
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&errMsg, IDS_VALID_VOLNAME_DOS);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatPascal:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (!DiskFSPascal::IsValidVolumeName(newNameA))
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&errMsg, IDS_VALID_VOLNAME_PASCAL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
case DiskImg::kFormatMacHFS:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
if (!DiskFSHFS::IsValidVolumeName(newNameA))
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
CheckedLoadString(&errMsg, IDS_VALID_VOLNAME_HFS);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
break;
default:
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg = L"Not supported by TestVolumeName!";
Add CheckedLoadString The static analyzer was annoyed that the return value from calls to CString::LoadString() was being ignored. This adds a wrapper function that checks the value and logs a failure message if the string can't be found.
2014-12-16 19:04:31 +00:00
break;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
return errMsg;
}
/*
* ===========================================================================
* DiskArchive -- set file properties
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
bool DiskArchive::SetProps(CWnd* pMsgWnd, GenericEntry* pGenericEntry,
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
const FileProps* pProps)
{
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*
* Technically we should reload the GenericArchive from the disk image,
* but the set of changes is pretty small, so we just make them here.
*/
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DIError dierr;
DiskEntry* pEntry = (DiskEntry*) pGenericEntry;
A2File* pFile = pEntry->GetA2File();
dierr = pFile->GetDiskFS()->SetFileInfo(pFile, pProps->fileType,
pProps->auxType, pProps->access);
if (dierr != kDIErrNone) {
CString errMsg;
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Unable to set file info: %hs.\n",
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
DiskImgLib::DIStrError(dierr));
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
return false;
}
/* do this in lieu of reloading GenericArchive */
pEntry->SetFileType(pFile->GetFileType());
pEntry->SetAuxType(pFile->GetAuxType());
pEntry->SetAccess(pFile->GetAccess());
/* DOS 3.2/3.3 may change these as well */
DiskImg::FSFormat fsFormat;
fsFormat = pFile->GetDiskFS()->GetDiskImg()->GetFSFormat();
if (fsFormat == DiskImg::kFormatDOS32 || fsFormat == DiskImg::kFormatDOS33) {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD(" (reloading additional fields after DOS SFI)");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pEntry->SetDataForkLen(pFile->GetDataLength());
pEntry->SetCompressedLen(pFile->GetDataSparseLength());
pEntry->SetSuspicious(pFile->GetQuality() == A2File::kQualitySuspicious);
}
/* clear the dirty flag in trivial cases */
(void) fpPrimaryDiskFS->Flush(DiskImg::kFlushFastOnly);
return true;
}
/*
* ===========================================================================
* DiskArchive -- transfer files to another archive
* ===========================================================================
*/
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
GenericArchive::XferStatus DiskArchive::XferSelection(CWnd* pMsgWnd,
SelectionSet* pSelSet, ActionProgressDialog* pActionProgress,
const XferFileOptions* pXferOpts)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
/*
* We get the open archive pointer and some options from "pXferOpts".
*
* The selection set was created with the "any" selection criteria, which
* means there's only one entry for each file regardless of whether it's
* forked or not.
*/
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("DiskArchive XferSelection!");
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
uint8_t* dataBuf = NULL;
uint8_t* rsrcBuf = NULL;
LocalFileDetails fileDetails;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
CString errMsg, extractErrMsg, cmpStr;
CString fixedPathName;
XferStatus retval = kXferFailed;
pXferOpts->fTarget->XferPrepare(pXferOpts);
SelectionEntry* pSelEntry = pSelSet->IterNext();
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
for ( ; pSelEntry != NULL; pSelEntry = pSelSet->IterNext()) {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
long dataLen=-1, rsrcLen=-1;
DiskEntry* pEntry = (DiskEntry*) pSelEntry->GetEntry();
int typeOverride = -1;
int result;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(dataBuf == NULL);
ASSERT(rsrcBuf == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (pEntry->GetDamaged()) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" XFER skipping damaged entry '%ls'",
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
(LPCWSTR) pEntry->GetDisplayName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
continue;
}
/*
* Do a quick de-colonizing pass for non-ProDOS volumes, then prepend
* the subvolume name (if any).
*/
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
fixedPathName = pEntry->GetPathNameUNI();
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (fixedPathName.IsEmpty())
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
fixedPathName = L"(no filename)";
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (pEntry->GetFSFormat() != DiskImg::kFormatProDOS)
fixedPathName.Replace(PathProposal::kDefaultStoredFssep, '.');
Improve filename handling when reading from archives This updates GenericEntry's filename handling to be more careful about Mac OS Roman vs. Unicode. Most of the work is still done with a CP-1252 conversion instead of MOR, but we now do a proper conversion on the "display name", so we see the right thing in the content list and file viewer. Copy & paste, disk-to-file-archive, and file-archive-to-disk conversions should work (correctly) as before. Extracted files will still have "_" or "%AA" instead of a Unicode TRADE MARK SIGN, but that's fine for now -- we can extract and re-add the files losslessly. The filenames are now stored in CStrings rather than WCHAR*. Also, fixed a bad initializer in the file-archive-to-disk conversion dialog.
2015-01-09 01:41:53 +00:00
if (!pEntry->GetSubVolName().IsEmpty()) {
CString tmpStr = pEntry->GetSubVolName();
tmpStr += (char) PathProposal::kDefaultStoredFssep;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
tmpStr += fixedPathName;
fixedPathName = tmpStr;
}
if (pEntry->GetRecordKind() == GenericEntry::kRecordKindVolumeDir) {
/* this is the volume dir */
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD(" XFER not transferring volume dir '%ls'",
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
(LPCWSTR) fixedPathName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
continue;
} else if (pEntry->GetRecordKind() == GenericEntry::kRecordKindDirectory) {
if (pXferOpts->fPreserveEmptyFolders) {
/* if this is an empty directory, create a fake entry */
cmpStr = fixedPathName;
cmpStr += (char)PathProposal::kDefaultStoredFssep;
if (pSelSet->CountMatchingPrefix(cmpStr) == 0) {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD("FOUND empty dir '%ls'", (LPCWSTR) fixedPathName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
cmpStr += kEmptyFolderMarker;
dataBuf = new unsigned char[1];
dataLen = 0;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
fileDetails.SetEntryKind(LocalFileDetails::kFileKindDataFork);
fileDetails.SetStrippedLocalPathName(cmpStr);
fileDetails.SetFileType(0); // NON
fileDetails.SetAccess(
pEntry->GetAccess() | GenericEntry::kAccessInvisible);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto have_stuff2;
} else {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD("NOT empty dir '%ls'", (LPCWSTR) fixedPathName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
}
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD(" XFER not transferring directory '%ls'",
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
(LPCWSTR) fixedPathName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
continue;
}
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" Xfer '%ls' (data=%d rsrc=%d)",
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
(LPCWSTR) fixedPathName, pEntry->GetHasDataFork(),
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pEntry->GetHasRsrcFork());
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
dataBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dataLen = 0;
result = pEntry->ExtractThreadToBuffer(GenericEntry::kDataThread,
(char**) &dataBuf, &dataLen, &extractErrMsg);
if (result == IDCANCEL) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Cancelled during data extract!");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail; /* abort anything that was pending */
} else if (result != IDOK) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Failed while extracting '%ls': %ls.",
Fix some issues identified by static analyzer Includes "printf format" annotation on debug log function.
2014-11-19 22:54:24 +00:00
(LPCWSTR) fixedPathName, (LPCWSTR) extractErrMsg);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
goto bail;
}
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(dataBuf != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ASSERT(dataLen >= 0);
#if 0
if (pXferOpts->fConvDOSText &&
DiskImg::UsesDOSFileStructure(pEntry->GetFSFormat()) &&
pEntry->GetFileType() == kFileTypeTXT)
{
/* don't need to convert EOL, so just strip in place */
long len;
unsigned char* ucp;
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" Converting DOS text in '%ls'", fixedPathName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
for (ucp = dataBuf, len = dataLen; len > 0; len--, ucp++)
*ucp = *ucp & 0x7f;
}
#endif
#if 0 // annoying to invoke PTX reformatter from here... ReformatHolder, etc.
if (pXferOpts->fConvPascalText &&
pEntry->GetFSFormat() == DiskImg::kFormatPascal &&
pEntry->GetFileType() == kFileTypePTX)
{
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("WOULD CONVERT ptx '%ls'", fixedPathName);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
#endif
if (pEntry->GetHasRsrcFork()) {
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
rsrcBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
rsrcLen = 0;
result = pEntry->ExtractThreadToBuffer(GenericEntry::kRsrcThread,
(char**) &rsrcBuf, &rsrcLen, &extractErrMsg);
if (result == IDCANCEL) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("Cancelled during rsrc extract!");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail; /* abort anything that was pending */
} else if (result != IDOK) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Failed while extracting '%ls': %ls.",
Fix some issues identified by static analyzer Includes "printf format" annotation on debug log function.
2014-11-19 22:54:24 +00:00
(LPCWSTR) fixedPathName, (LPCWSTR) extractErrMsg);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
goto bail;
}
} else {
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(rsrcBuf == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
if (pEntry->GetHasDataFork() && pEntry->GetHasRsrcFork()) {
fileDetails.SetEntryKind(LocalFileDetails::kFileKindBothForks);
} else if (pEntry->GetHasDataFork()) {
fileDetails.SetEntryKind(LocalFileDetails::kFileKindDataFork);
} else if (pEntry->GetHasRsrcFork()) {
fileDetails.SetEntryKind(LocalFileDetails::kFileKindRsrcFork);
} else {
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ASSERT(false);
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
fileDetails.SetEntryKind(LocalFileDetails::kFileKindUnknown);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
/*
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
* Set up the rest of the LocalFileDetails fields.
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
*/
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
fileDetails.SetStrippedLocalPathName(fixedPathName);
fileDetails.SetFileType(pEntry->GetFileType());
fileDetails.SetAccess(pEntry->GetAccess());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
have_stuff2:
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
fileDetails.SetFileSysFmt(pEntry->GetSourceFS());
fileDetails.SetFssep(PathProposal::kDefaultStoredFssep);
fileDetails.SetExtraType(pEntry->GetAuxType());
fileDetails.SetStorageType(kNuStorageUnknown); // let NufxLib deal
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
NuDateTime ndt;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
time_t when;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
when = time(NULL);
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
UNIXTimeToDateTime(&when, &ndt);
fileDetails.SetArchiveWhen(ndt);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
when = pEntry->GetModWhen();
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
UNIXTimeToDateTime(&when, &ndt);
fileDetails.SetModWhen(ndt);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
when = pEntry->GetCreateWhen();
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
UNIXTimeToDateTime(&when, &ndt);
fileDetails.SetCreateWhen(ndt);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
pActionProgress->SetArcName(fileDetails.GetStrippedLocalPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (pActionProgress->SetProgress(0) == IDCANCEL) {
retval = kXferCancelled;
goto bail;
}
errMsg = pXferOpts->fTarget->XferFile(&fileDetails, &dataBuf, dataLen,
&rsrcBuf, rsrcLen);
if (!errMsg.IsEmpty()) {
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("XferFile failed!");
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"Failed while transferring '%ls': %ls.",
(LPCWSTR) pEntry->GetDisplayName(), (LPCWSTR) errMsg);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
ShowFailureMsg(pMsgWnd, errMsg, IDS_FAILED);
goto bail;
}
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(dataBuf == NULL);
ASSERT(rsrcBuf == NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
if (pActionProgress->SetProgress(100) == IDCANCEL) {
retval = kXferCancelled;
goto bail;
}
}
//MainWindow* pMainWin;
//pMainWin = (MainWindow*)::AfxGetMainWnd();
//pMainWin->EventPause(1000);
retval = kXferOK;
bail:
if (retval != kXferOK)
pXferOpts->fTarget->XferAbort(pMsgWnd);
else
pXferOpts->fTarget->XferFinish(pMsgWnd);
delete[] dataBuf;
delete[] rsrcBuf;
return retval;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
void DiskArchive::XferPrepare(const XferFileOptions* pXferOpts)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("DiskArchive::XferPrepare");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
//fpPrimaryDiskFS->SetParameter(DiskFS::kParmProDOS_AllowLowerCase,
// pXferOpts->fAllowLowerCase);
//fpPrimaryDiskFS->SetParameter(DiskFS::kParmProDOS_AllocSparse,
// pXferOpts->fUseSparseBlocks);
fpPrimaryDiskFS->SetParameter(DiskFS::kParm_CreateUnique, true);
//fXferStoragePrefix = pXferOpts->fStoragePrefix;
fpXferTargetFS = pXferOpts->fpTargetFS;
}
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
CString DiskArchive::XferFile(LocalFileDetails* pDetails, uint8_t** pDataBuf,
Use types with explicit sizes Much of what the "reformat" code does involves processing data that is 8, 16, or 32 bits. We want to use size-specific types from stdint.h (e.g. uint16_t) rather than "unsigned short". This was a quick pass to replace the various "unsigned" declarations. More can be done here and elsewhere.
2014-11-21 02:10:18 +00:00
long dataLen, uint8_t** pRsrcBuf, long rsrcLen)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
DiskFS* pDiskFS;
CString errMsg;
DIError dierr = kDIErrNone;
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI(" XFER: transfer '%ls' (dataLen=%ld rsrcLen=%ld)",
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
(LPCWSTR) pDetails->GetStrippedLocalPathName(), dataLen, rsrcLen);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
ASSERT(pDataBuf != NULL);
ASSERT(pRsrcBuf != NULL);
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
const DiskFS::CreateParms& createParms = pDetails->GetCreateParms();
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
if (fpXferTargetFS == NULL)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
pDiskFS = fpPrimaryDiskFS;
else
pDiskFS = fpXferTargetFS;
/*
* Strip the high ASCII from DOS and RDOS text files, unless we're adding
* them to a DOS disk. Likewise, if we're adding non-DOS text files to
* a DOS disk, we need to add the high bit.
*
* DOS converts both TXT and SRC to 'T', so we have to handle both here.
* Ideally we'd just ask DOS, "do you think this is a text file?", but it's
* not worth adding a new interface just for that.
*/
bool srcIsDOS, dstIsDOS;
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
srcIsDOS = DiskImg::UsesDOSFileStructure(pDetails->GetFileSysFmt());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
dstIsDOS = DiskImg::UsesDOSFileStructure(pDiskFS->GetDiskImg()->GetFSFormat());
if (dataLen > 0 &&
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
(pDetails->GetFileType() == kFileTypeTXT ||
pDetails->GetFileType() == kFileTypeSRC))
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
unsigned char* ucp = *pDataBuf;
long len = dataLen;
if (srcIsDOS && !dstIsDOS) {
Fix some issues identified by static analyzer Includes "printf format" annotation on debug log function.
2014-11-19 22:54:24 +00:00
LOGD(" Stripping high ASCII from '%ls'",
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
(LPCWSTR) pDetails->GetStrippedLocalPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
while (len--)
*ucp++ &= 0x7f;
} else if (!srcIsDOS && dstIsDOS) {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD(" Adding high ASCII to '%ls'",
(LPCWSTR) pDetails->GetStrippedLocalPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
while (len--) {
if (*ucp != '\0')
*ucp |= 0x80;
ucp++;
}
} else if (srcIsDOS && dstIsDOS) {
Fix some issues identified by static analyzer Includes "printf format" annotation on debug log function.
2014-11-19 22:54:24 +00:00
LOGD(" --- not altering DOS-to-DOS text '%ls'",
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
(LPCWSTR) pDetails->GetStrippedLocalPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
} else {
Improve filename handling when adding files Most of this change is a conversion of the old FileDetails struct into a new LocalFileDetails class. The new class keeps the members private, and keeps the Unicode and MOR representations of the string separate. The NuFX and DiskImg libraries don't support UTF-16 filenames, so we stil can't add files with non-CP-1252 filenames, but we're a step closer. Also, update NufxLib with a couple of fixes from the main project. Also, fix handling of "%00" when adding files. Also, mark most of the A2FileDOS fields private. Not sure why they weren't.
2015-01-05 05:04:01 +00:00
LOGD(" --- non-DOS transfer '%ls'",
(LPCWSTR) pDetails->GetStrippedLocalPathName());
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
}
}
/* add a file with one or two forks */
if (createParms.storageType == kNuStorageDirectory) {
ASSERT(dataLen < 0 && rsrcLen < 0);
} else {
ASSERT(dataLen >= 0 || rsrcLen >= 0); // at least one fork
}
/* if we still have something to write, write it */
dierr = AddForksToDisk(pDiskFS, &createParms, *pDataBuf, dataLen,
*pRsrcBuf, rsrcLen);
if (dierr != kDIErrNone) {
Large set of changes to restore CiderPress build. CiderPress and MDC now compile, and execute far enough to open their respective "about" boxes, but I doubt they'll do much more than that. * Switch from MBCS to UNICODE APIs Microsoft switched to UTF-16 (by way of UCS-2) a long time ago, and the support for MBCS seems to be getting phased out. So it's time to switch to wide strings. This is a bit awkward for CiderPress because it works with disk and file archives with 8-bit filenames, and I want NufxLib and DiskImgLib to continue to work on Linux (which has largely taken the UTF-8 approach to Unicode). The libraries will continue to work with 8-bit filenames, with CiderPress/MDC doing the conversion at the appropriate point. There were a couple of places where strings from a structure handed back by one of the libraries were used directly in the UI, or vice-versa, which is a problem because we have nowhere to store the result of the conversion. These currently have fixed place-holder "xyzzy" strings. All UI strings are now wide. Various format strings now use "%ls" and "%hs" to explicitly specify wide and narrow. This doesn't play well with gcc, so only the Windows-specific parts use those. * Various updates to vcxproj files The project-file conversion had some cruft that is now largely gone. The build now has a common output directory for the EXEs and libraries, avoiding the old post-build copy steps. * Added zlib 1.2.8 and nufxlib 2.2.2 source snapshots The old "prebuilts" directory is now gone. The libraries are now built as part of building the apps. I added a minimal set of files for zlib, and a full set for nufxlib. The Linux-specific nufxlib goodies are included for the benefit of the Linux utilities, which are currently broken (don't build). * Replace symbols used for include guards Symbols with a leading "__" are reserved.
2014-11-10 23:32:55 +00:00
errMsg.Format(L"%hs", DiskImgLib::DIStrError(dierr));
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
goto bail;
}
/* clean up */
delete[] *pDataBuf;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
*pDataBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
delete[] *pRsrcBuf;
Change "nil" to "NULL"
2014-11-18 05:13:13 +00:00
*pRsrcBuf = NULL;
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
bail:
return errMsg;
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
void DiskArchive::XferAbort(CWnd* pMsgWnd)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
// Can't undo previous actions.
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("DiskArchive::XferAbort");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
InternalReload(pMsgWnd);
}
Relocate method comments This moves method comments from the .cpp file to the .h file, where users of the methods can find them. This also makes it possible for the IDE to show the comments when you mouse-hover over the method name, though Visual Studio is a bit weak in this regard. Also, added "override" keywords on overridden methods. Reasonably current versions of popular compilers seem to support this. Also, don't have the return type on a separate line in the .cpp file. The motivation for the practice -- quickly finding a method definition with "^name" -- is less useful in C++ than C, and modern IDEs provide more convenient ways to do the same thing. Also, do some more conversion from unsigned types to uintXX_t. This commit is primarily for the "app" directory.
2014-11-21 21:18:20 +00:00
void DiskArchive::XferFinish(CWnd* pMsgWnd)
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
{
Convert WSMG to LOG Mostly a bulk conversion of debug messages, primarily with sed: sed -e 's/\(WMSG[0-9]\)\(.*\)\(\\n"\)/LOGI\2"/' This removes the '\n' from the end of the log messages, and sets them all to "info" severity. We want to prefix each line with file/line and/or a timestamp, so it doesn't make sense to have a partial line, and there's no value in embedding the '\n' in every string.
2014-11-18 21:05:15 +00:00
LOGI("DiskArchive::XferFinish");
Normalize indentation and EOL This updates all source files to use spaces instead of tabs for indentation. It also normalizes the end-of-line markers to be Windows-style CRLF, and ensures that all files end with EOL. No substantive changes were made; "diff -w" is empty.
2014-11-04 00:26:53 +00:00
InternalReload(pMsgWnd);
}
Reference in New Issue Copy Permalink