hush/libbb/get_line_from_file.c

151 lines
3.6 KiB
C
Raw Normal View History

/* vi: set sw=4 ts=4: */
/*
* Utility routines.
*
* Copyright (C) 2005, 2006 Rob Landley <rob@landley.net>
* Copyright (C) 2004 Erik Andersen <andersen@codepoet.org>
* Copyright (C) 2001 Matt Krai
*
* Licensed under GPLv2 or later, see file LICENSE in this tarball for details.
*/
#include "libbb.h"
/* This function reads an entire line from a text file, up to a newline
* or NUL byte, inclusive. It returns a malloc'ed char * which
* must be free'ed by the caller. If end is NULL '\n' isn't considered
* end of line. If end isn't NULL, length of the chunk read is stored in it.
* Return NULL if EOF/error */
char* FAST_FUNC bb_get_chunk_from_file(FILE *file, int *end)
{
int ch;
int idx = 0;
char *linebuf = NULL;
int linebufsz = 0;
2003-03-19 09:13:01 +00:00
while ((ch = getc(file)) != EOF) {
/* grow the line buffer as necessary */
2006-12-02 17:58:10 +00:00
if (idx >= linebufsz) {
2007-04-12 00:32:05 +00:00
linebufsz += 80;
linebuf = xrealloc(linebuf, linebufsz);
2003-03-19 09:13:01 +00:00
}
linebuf[idx++] = (char) ch;
if (!ch || (end && ch == '\n'))
break;
}
if (end)
*end = idx;
2003-03-19 09:13:01 +00:00
if (linebuf) {
2006-12-02 17:58:10 +00:00
// huh, does fgets discard prior data on error like this?
// I don't think so....
//if (ferror(file)) {
// free(linebuf);
// return NULL;
//}
linebuf = xrealloc(linebuf, idx + 1);
2006-12-02 17:58:10 +00:00
linebuf[idx] = '\0';
2003-03-19 09:13:01 +00:00
}
return linebuf;
}
/* Get line, including trailing \n if any */
char* FAST_FUNC xmalloc_fgets(FILE *file)
2003-03-19 09:13:01 +00:00
{
int i;
return bb_get_chunk_from_file(file, &i);
2003-03-19 09:13:01 +00:00
}
/* Get line. Remove trailing \n */
char* FAST_FUNC xmalloc_fgetline(FILE *file)
2003-03-19 09:13:01 +00:00
{
int i;
char *c = bb_get_chunk_from_file(file, &i);
if (i && c[--i] == '\n')
c[i] = '\0';
return c;
}
#if 0
/* Faster routines (~twice as fast). +170 bytes. Unused as of 2008-07.
*
* NB: they stop at NUL byte too.
* Performance is important here. Think "grep 50gigabyte_file"...
* Ironically, grep can't use it because of NUL issue.
* We sorely need C lib to provide fgets which reports size!
*
* Update:
* Actually, uclibc and glibc have it. man getline. It's GNUism,
* but very useful one (if it's as fast as this code).
* TODO:
* - currently, sed and sort use bb_get_chunk_from_file and heavily
* depend on its "stop on \n or \0" behavior, and STILL they fail
* to handle all cases with embedded NULs correctly. So:
* - audit sed and sort; convert them to getline FIRST.
* - THEN ditch bb_get_chunk_from_file, replace it with getline.
* - provide getline implementation for non-GNU systems.
*/
static char* xmalloc_fgets_internal(FILE *file, int *sizep)
{
int len;
int idx = 0;
char *linebuf = NULL;
while (1) {
char *r;
linebuf = xrealloc(linebuf, idx + 0x100);
r = fgets(&linebuf[idx], 0x100, file);
if (!r) {
/* need to terminate in case this is error
* (EOF puts NUL itself) */
linebuf[idx] = '\0';
break;
}
/* stupid. fgets knows the len, it should report it somehow */
len = strlen(&linebuf[idx]);
idx += len;
if (len != 0xff || linebuf[idx - 1] == '\n')
break;
}
*sizep = idx;
if (idx) {
/* xrealloc(linebuf, idx + 1) is up to caller */
return linebuf;
}
free(linebuf);
return NULL;
}
/* Get line, remove trailing \n */
char* FAST_FUNC xmalloc_fgetline_fast(FILE *file)
{
int sz;
char *r = xmalloc_fgets_internal(file, &sz);
if (r && r[sz - 1] == '\n')
r[--sz] = '\0';
return r; /* not xrealloc(r, sz + 1)! */
}
char* FAST_FUNC xmalloc_fgets(FILE *file)
{
int sz;
return xmalloc_fgets_internal(file, &sz);
}
/* Get line, remove trailing \n */
char* FAST_FUNC xmalloc_fgetline(FILE *file)
{
int sz;
char *r = xmalloc_fgets_internal(file, &sz);
if (!r)
return r;
if (r[sz - 1] == '\n')
r[--sz] = '\0';
return xrealloc(r, sz + 1);
}
#endif