contiki/apps/webbrowser/htmlparser.c

813 lines
22 KiB
C
Raw Normal View History

/*
* Copyright (c) 2002, Adam Dunkels.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following
* disclaimer in the documentation and/or other materials provided
* with the distribution.
* 3. The name of the author may not be used to endorse or promote
* products derived from this software without specific prior
* written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS
* OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
* WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY
* DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
* GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
* WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
* NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* This file is part of the Contiki desktop environment
*
*
*/
/* htmlparser.c:
*
* Implements a very simplistic HTML parser. It recognizes HTML links
* (<a href>-tags), HTML img alt tags, a few text flow break tags
G * (<br>, <p>, <h>), the <li> tag (but does not even try to
* distinguish between <ol> or <ul>) as well as HTML comment tags
* (<!-- -->).
*
* To save memory, the HTML parser is state machine driver, which
* means that it will shave off one character from the HTML page,
* process that character, and return to the next. Another way of
* doing it would be to buffer a number of characters and process them
* together.
*
* The main function in this file is the htmlparser_parse() function
* which takes a htmlparser_state structur and a part of an HTML file
* as an argument. The htmlparser_parse() function will call the
* helper functions parse_char() and parse_tag(). Those functions will
* in turn call the two callback functions htmlparser_char() and
* htmlparser_tag(). Those functions must be implemented by the using
* module (e.g., a web browser program).
*
* htmlparser_char() will be called for every non-tag character.
*
* htmlparser_tag() will be called whenever a full tag has been found.
*
*/
#include <string.h>
#include "contiki.h"
#include "html-strings.h"
#include "www.h"
#include "htmlparser.h"
#if 1
#define PRINTF(x)
#else
#include <stdio.h>
#define PRINTF(x) printf x
#endif
/*-----------------------------------------------------------------------------------*/
#define ISO_A 0x41
#define ISO_B 0x42
#define ISO_E 0x45
#define ISO_F 0x46
#define ISO_G 0x47
#define ISO_H 0x48
#define ISO_I 0x49
#define ISO_L 0x4c
#define ISO_M 0x4d
#define ISO_P 0x50
#define ISO_R 0x52
#define ISO_T 0x54
#define ISO_a (ISO_A | 0x20)
#define ISO_b (ISO_B | 0x20)
#define ISO_e (ISO_E | 0x20)
#define ISO_f (ISO_F | 0x20)
#define ISO_g (ISO_G | 0x20)
#define ISO_h (ISO_H | 0x20)
#define ISO_i (ISO_I | 0x20)
#define ISO_l (ISO_L | 0x20)
#define ISO_m (ISO_M | 0x20)
#define ISO_p (ISO_P | 0x20)
#define ISO_r (ISO_R | 0x20)
#define ISO_t (ISO_T | 0x20)
#define ISO_ht 0x09
#define ISO_nl 0x0a
#define ISO_cr 0x0d
#define ISO_space 0x20
#define ISO_bang 0x21
#define ISO_citation 0x22
#define ISO_ampersand 0x26
#define ISO_citation2 0x27
#define ISO_asterisk 0x2a
#define ISO_dash 0x2d
#define ISO_slash 0x2f
#define ISO_semicolon 0x3b
#define ISO_lt 0x3c
#define ISO_eq 0x3d
#define ISO_gt 0x3e
#define MINORSTATE_NONE 0
#define MINORSTATE_TEXT 1 /* Parse normal text */
#define MINORSTATE_EXTCHAR 2 /* Check for semi-colon */
#define MINORSTATE_TAG 3 /* Check for name of tag. */
#define MINORSTATE_TAGEND 4 /* Scan for end of tag. */
#define MINORSTATE_TAGATTR 5 /* Parse tag attr. */
#define MINORSTATE_TAGATTRSPACE 6 /* Parse optional space after tag
attr. */
#define MINORSTATE_TAGATTRPARAM 7 /* Parse tag attr parameter. */
#define MINORSTATE_TAGATTRPARAMNQ 8 /* Parse tag attr parameter without
quotation marks. */
#define MINORSTATE_HTMLCOMMENT 9 /* Scan for HTML comment end */
#define MAJORSTATE_NONE 0
#define MAJORSTATE_BODY 1
#define MAJORSTATE_LINK 2
#define MAJORSTATE_FORM 3
#define MAJORSTATE_DISCARD 4
#define MAJORSTATE_SCRIPT 5
struct htmlparser_state {
unsigned char minorstate;
char tag[20];
unsigned char tagptr;
char tagattr[20];
unsigned char tagattrptr;
char tagattrparam[WWW_CONF_MAX_URLLEN + 1];
unsigned char tagattrparamptr;
2015-06-05 13:11:57 +00:00
unsigned char quotechar;
unsigned char majorstate, lastmajorstate;
char linkurl[WWW_CONF_MAX_URLLEN + 1];
char word[WWW_CONF_WEBPAGE_WIDTH];
unsigned char wordlen;
#if WWW_CONF_FORMS
char formaction[WWW_CONF_MAX_FORMACTIONLEN + 1];
unsigned char inputtype;
char inputname[WWW_CONF_MAX_INPUTNAMELEN + 1];
char inputvalue[WWW_CONF_MAX_INPUTVALUELEN + 1];
unsigned char inputvaluesize;
#endif /* WWW_CONF_FORMS */
};
static struct htmlparser_state s;
/*-----------------------------------------------------------------------------------*/
2007-11-18 01:41:00 +00:00
static char last[1] = {(char)0xff};
static const char *tags[] = {
#define TAG_FIRST 0
#define TAG_SLASHA 0
html_slasha,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_SLASHDIV 1
html_slashdiv,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_SLASHFORM 2
html_slashform,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_SLASHH 3
html_slashh,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_SLASHSCRIPT 4
html_slashscript,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_SLASHSELECT 5
html_slashselect,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_SLASHSTYLE 6
html_slashstyle,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_A 7
html_a,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_BODY 8
html_body,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_BR 9
html_br,
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#define TAG_FORM 10
html_form,
#define TAG_H1 11
html_h1,
#define TAG_H2 12
html_h2,
#define TAG_H3 13
html_h3,
#define TAG_H4 14
html_h4,
#define TAG_IMG 15
html_img,
#define TAG_INPUT 16
html_input,
#define TAG_LI 17
html_li,
#define TAG_P 18
html_p,
#define TAG_SCRIPT 19
html_script,
#define TAG_SELECT 20
html_select,
#define TAG_STYLE 21
html_style,
#define TAG_TR 22
html_tr,
#define TAG_LAST 23
last,
};
/*-----------------------------------------------------------------------------------*/
static unsigned char
iswhitespace(char c)
{
return (c == ISO_space ||
c == ISO_nl ||
c == ISO_cr ||
c == ISO_ht);
}
/*-----------------------------------------------------------------------------------*/
#if WWW_CONF_FORMS
static void
init_input(void)
{
s.inputtype = HTMLPARSER_INPUTTYPE_NONE;
s.inputname[0] = s.inputvalue[0] =
s.formaction[WWW_CONF_MAX_FORMACTIONLEN] =
s.inputname[WWW_CONF_MAX_INPUTNAMELEN] =
s.inputvalue[WWW_CONF_MAX_INPUTVALUELEN] = 0;
s.inputvaluesize = 20; /* De facto default size */
}
#endif /* WWW_CONF_FORMS */
/*-----------------------------------------------------------------------------------*/
void
htmlparser_init(void)
{
s.majorstate = s.lastmajorstate = MAJORSTATE_DISCARD;
s.minorstate = MINORSTATE_TEXT;
2015-06-05 13:11:57 +00:00
s.wordlen = 0;
#if WWW_CONF_FORMS
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
s.formaction[0] = 0;
#endif /* WWW_CONF_FORMS */
}
/*-----------------------------------------------------------------------------------*/
static char
lowercase(char c)
{
/* XXX: This is a *brute force* approach to lower-case
converting and should *not* be used anywhere else! It
works for our purposes, however (i.e., HTML tags). */
if(c > 0x40) {
return (c & 0x1f) | 0x60;
} else {
return c;
}
}
/*-----------------------------------------------------------------------------------*/
static void
endtagfound(void)
{
s.tag[s.tagptr] = 0;
s.tagattr[s.tagattrptr] = 0;
s.tagattrparam[s.tagattrparamptr] = 0;
}
/*-----------------------------------------------------------------------------------*/
static void
switch_majorstate(unsigned char newstate)
{
if(s.majorstate != newstate) {
PRINTF(("Switching state from %d to %d (%d)\n", s.majorstate, newstate, s.lastmajorstate));
s.lastmajorstate = s.majorstate;
s.majorstate = newstate;
}
}
/*-----------------------------------------------------------------------------------*/
static void
add_char(unsigned char c)
{
if(s.wordlen < WWW_CONF_WEBPAGE_WIDTH - 1 && c < 0x80) {
s.word[s.wordlen] = c;
++s.wordlen;
}
}
/*-----------------------------------------------------------------------------------*/
static void
do_word(void)
{
if(s.wordlen > 0) {
if(s.majorstate == MAJORSTATE_LINK) {
if(s.word[s.wordlen - 1] != ISO_space) {
add_char(ISO_space);
}
} else if(s.majorstate >= MAJORSTATE_DISCARD) {
s.wordlen = 0;
} else {
s.word[s.wordlen] = '\0';
htmlparser_word(s.word, s.wordlen);
s.wordlen = 0;
}
}
}
/*-----------------------------------------------------------------------------------*/
static void
newline(void)
{
do_word();
htmlparser_newline();
}
/*-----------------------------------------------------------------------------------*/
static unsigned char
find_tag(char *tag)
{
static unsigned char first, last, i, tabi;
static char tagc;
first = TAG_FIRST;
last = TAG_LAST;
i = 0;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
do {
tagc = tag[i];
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
if((tagc == 0 || tagc == ISO_slash) && tags[first][i] == 0) {
return first;
}
tabi = first;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
/* First, find first matching tag from table. */
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
while(tagc > (tags[tabi])[i] && tabi < last) {
++tabi;
}
first = tabi;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
/* Second, find last matching tag from table. */
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
while(tagc == (tags[tabi])[i] && tabi < last) {
++tabi;
}
last = tabi;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
/* If first and last matching tags are equal, we have a non-match
and return. Else we continue with the next character. */
++i;
} while(last != first);
return TAG_LAST;
}
/*-----------------------------------------------------------------------------------*/
static void
parse_tag(void)
{
static char *tagattrparam;
static unsigned char tag;
2006-10-06 21:28:18 +00:00
static unsigned char size;
tag = find_tag(s.tag);
/* If we are inside a <script> we mustn't interpret any tags
(inside JavaScript strings) but wait for the </script>. */
if(s.majorstate == MAJORSTATE_SCRIPT && tag != TAG_SLASHSCRIPT) {
return;
}
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
PRINTF(("Parsing tag '%s' '%s' '%s'\n", s.tag, s.tagattr, s.tagattrparam));
switch(tag) {
case TAG_P:
case TAG_H1:
case TAG_H2:
case TAG_H3:
case TAG_H4:
newline();
/* FALLTHROUGH */
case TAG_BR:
case TAG_TR:
case TAG_SLASHDIV:
case TAG_SLASHH:
newline();
break;
case TAG_LI:
if(s.tagattr[0] == 0) {
newline();
add_char(ISO_asterisk);
add_char(ISO_space);
}
break;
case TAG_SCRIPT:
switch_majorstate(MAJORSTATE_SCRIPT);
break;
case TAG_STYLE:
case TAG_SELECT:
switch_majorstate(MAJORSTATE_DISCARD);
break;
case TAG_SLASHSCRIPT:
case TAG_SLASHSTYLE:
case TAG_SLASHSELECT:
do_word();
switch_majorstate(s.lastmajorstate);
break;
case TAG_BODY:
s.majorstate = s.lastmajorstate = MAJORSTATE_BODY;
break;
case TAG_IMG:
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
if(strncmp(s.tagattr, html_alt, sizeof(html_alt)) == 0 && s.tagattrparam[0] != 0) {
add_char(ISO_lt);
tagattrparam = &s.tagattrparam[0];
while(*tagattrparam) {
add_char(*tagattrparam);
++tagattrparam;
}
add_char(ISO_gt);
do_word();
}
break;
case TAG_A:
PRINTF(("A %s %s\n", s.tagattr, s.tagattrparam));
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
if(strncmp(s.tagattr, html_href, sizeof(html_href)) == 0 && s.tagattrparam[0] != 0) {
strcpy(s.linkurl, s.tagattrparam);
do_word();
switch_majorstate(MAJORSTATE_LINK);
}
break;
case TAG_SLASHA:
if(s.majorstate == MAJORSTATE_LINK) {
switch_majorstate(s.lastmajorstate);
s.word[s.wordlen] = 0;
htmlparser_link(s.word, s.wordlen, s.linkurl);
s.wordlen = 0;
}
break;
#if WWW_CONF_FORMS
case TAG_FORM:
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
/* First check if we are called at the end of a form tag. If
so, we should propagate the form action. */
if(s.tagattr[0] == 0 && s.formaction[0] != 0) {
htmlparser_form(s.formaction);
init_input();
} else {
PRINTF(("Form tag\n"));
switch_majorstate(MAJORSTATE_FORM);
if(strncmp(s.tagattr, html_action, sizeof(html_action)) == 0) {
PRINTF(("Form action '%s'\n", s.tagattrparam));
strncpy(s.formaction, s.tagattrparam, WWW_CONF_MAX_FORMACTIONLEN - 1);
}
}
break;
case TAG_SLASHFORM:
switch_majorstate(MAJORSTATE_BODY);
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
s.formaction[0] = 0;
break;
case TAG_INPUT:
if(s.majorstate == MAJORSTATE_FORM) {
/* First check if we are called at the end of an input tag. If
so, we should render the input widget. */
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
if(s.tagattr[0] == 0 && s.inputname[0] != 0) {
PRINTF(("Render input type %d\n", s.inputtype));
switch(s.inputtype) {
case HTMLPARSER_INPUTTYPE_NONE:
case HTMLPARSER_INPUTTYPE_TEXT:
case HTMLPARSER_INPUTTYPE_HIDDEN:
htmlparser_inputfield(s.inputtype, s.inputvaluesize, s.inputvalue, s.inputname);
break;
case HTMLPARSER_INPUTTYPE_SUBMIT:
case HTMLPARSER_INPUTTYPE_IMAGE:
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
htmlparser_submitbutton(s.inputvalue, s.inputname);
break;
}
init_input();
} else {
PRINTF(("Input '%s' '%s'\n", s.tagattr, s.tagattrparam));
if(strncmp(s.tagattr, html_type, sizeof(html_type)) == 0) {
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
if(strncmp(s.tagattrparam, html_submit, sizeof(html_submit)) == 0) {
s.inputtype = HTMLPARSER_INPUTTYPE_SUBMIT;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
} else if(strncmp(s.tagattrparam, html_image, sizeof(html_image)) == 0) {
s.inputtype = HTMLPARSER_INPUTTYPE_IMAGE;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
} else if(strncmp(s.tagattrparam, html_text, sizeof(html_text)) == 0) {
s.inputtype = HTMLPARSER_INPUTTYPE_TEXT;
} else if(strncmp(s.tagattrparam, html_hidden, sizeof(html_hidden)) == 0) {
s.inputtype = HTMLPARSER_INPUTTYPE_HIDDEN;
} else {
s.inputtype = HTMLPARSER_INPUTTYPE_OTHER;
}
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
} else if(strncmp(s.tagattr, html_name, sizeof(html_name)) == 0) {
strncpy(s.inputname, s.tagattrparam, WWW_CONF_MAX_INPUTNAMELEN);
} else if(strncmp(s.tagattr, html_alt, sizeof(html_alt)) == 0 &&
s.inputtype == HTMLPARSER_INPUTTYPE_IMAGE) {
strncpy(s.inputvalue, s.tagattrparam, WWW_CONF_MAX_INPUTVALUELEN);
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
} else if(strncmp(s.tagattr, html_value, sizeof(html_value)) == 0) {
strncpy(s.inputvalue, s.tagattrparam, WWW_CONF_MAX_INPUTVALUELEN);
} else if(strncmp(s.tagattr, html_size, sizeof(html_size)) == 0) {
size = 0;
if(s.tagattrparam[0] >= '0' &&
s.tagattrparam[0] <= '9') {
size = s.tagattrparam[0] - '0';
if(s.tagattrparam[1] >= '0' &&
s.tagattrparam[1] <= '9') {
size = size * 10 + (s.tagattrparam[1] - '0');
}
}
if(size >= WWW_CONF_MAX_INPUTVALUELEN) {
size = WWW_CONF_MAX_INPUTVALUELEN - 1;
}
s.inputvaluesize = size;
}
}
}
break;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
#endif /* WWW_CONF_FORMS */
}
}
/*-----------------------------------------------------------------------------------*/
static uint16_t
parse_word(char *data, uint8_t dlen)
{
static uint8_t i;
static uint8_t len;
unsigned char c;
len = dlen;
switch(s.minorstate) {
case MINORSTATE_TEXT:
for(i = 0; i < len; ++i) {
c = data[i];
if(iswhitespace(c)) {
do_word();
} else if(c == ISO_lt) {
s.minorstate = MINORSTATE_TAG;
s.tagptr = 0;
break;
} else if(c == ISO_ampersand) {
s.minorstate = MINORSTATE_EXTCHAR;
break;
} else {
add_char(c);
}
}
break;
case MINORSTATE_EXTCHAR:
for(i = 0; i < len; ++i) {
c = data[i];
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
if(c == ISO_semicolon) {
s.minorstate = MINORSTATE_TEXT;
add_char(' ');
break;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
} else if(iswhitespace(c)) {
s.minorstate = MINORSTATE_TEXT;
add_char('&');
add_char(' ');
break;
}
}
break;
case MINORSTATE_TAG:
/* If we are inside a <srcipt> we mustn't mistake a JavaScript
equation with a '<' as a tag. So we check for the very next
character to be a '/' as we're only interested in parsing
the </script>. */
if(s.majorstate == MAJORSTATE_SCRIPT && data[0] != ISO_slash) {
s.minorstate = MINORSTATE_TEXT;
break;
}
/* We are currently parsing within the name of a tag. We check
for the end of a tag (the '>' character) or whitespace (which
indicates that we should parse a tag attr argument
instead). */
for(i = 0; i < len; ++i) {
c = data[i];
if(c == ISO_gt) {
/* Full tag found. We continue parsing regular text. */
s.minorstate = MINORSTATE_TEXT;
s.tagattrptr = s.tagattrparamptr = 0;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
endtagfound();
parse_tag();
break;
} else if(iswhitespace(c)) {
/* The name of the tag found. We continue parsing the tag
attr.*/
s.minorstate = MINORSTATE_TAGATTR;
s.tagattrptr = 0;
endtagfound();
break;
} else {
/* Keep track of the name of the tag, but convert it to
lower case. */
s.tag[s.tagptr] = lowercase(c);
++s.tagptr;
/* Check if the ->tag field is full. If so, we just eat up
any data left in the tag. */
if(s.tagptr == sizeof(s.tag)) {
s.minorstate = MINORSTATE_TAGEND;
break;
}
}
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
/* Check for HTML comment, indicated by <!-- */
if(s.tagptr == 3 &&
s.tag[0] == ISO_bang &&
s.tag[1] == ISO_dash &&
s.tag[2] == ISO_dash) {
PRINTF(("Starting comment...\n"));
s.minorstate = MINORSTATE_HTMLCOMMENT;
s.tagptr = 0;
endtagfound();
break;
}
}
break;
case MINORSTATE_TAGATTR:
/* We parse the "tag attr", i.e., the "href" in <a
href="...">. */
for(i = 0; i < len; ++i) {
c = data[i];
if(c == ISO_gt) {
/* Full tag found. */
s.minorstate = MINORSTATE_TEXT;
s.tagattrparamptr = 0;
s.tagattrptr = 0;
endtagfound();
parse_tag();
s.tagptr = 0;
endtagfound();
break;
} else if(iswhitespace(c)) {
if(s.tagattrptr == 0) {
/* Discard leading spaces. */
} else {
/* A non-leading space is the end of the attribute. */
s.tagattrparamptr = 0;
endtagfound();
parse_tag();
s.minorstate = MINORSTATE_TAGATTRSPACE;
break;
}
} else if(c == ISO_eq) {
s.minorstate = MINORSTATE_TAGATTRPARAMNQ;
s.tagattrparamptr = 0;
endtagfound();
break;
} else {
s.tagattr[s.tagattrptr] = lowercase(c);
++s.tagattrptr;
/* Check if the "tagattr" field is full. If so, we just eat
up any data left in the tag. */
if(s.tagattrptr == sizeof(s.tagattr)) {
s.minorstate = MINORSTATE_TAGEND;
break;
}
}
}
break;
case MINORSTATE_TAGATTRSPACE:
for(i = 0; i < len; ++i) {
c = data[i];
if(iswhitespace(c)) {
/* Discard spaces. */
} else if(c == ISO_eq) {
s.minorstate = MINORSTATE_TAGATTRPARAMNQ;
s.tagattrparamptr = 0;
endtagfound();
parse_tag();
break;
} else {
s.tagattr[0] = lowercase(c);
s.tagattrptr = 1;
s.minorstate = MINORSTATE_TAGATTR;
break;
}
}
break;
case MINORSTATE_TAGATTRPARAMNQ:
/* We are parsing the "tag attr parameter", i.e., the link part
in <a href="link">. */
for(i = 0; i < len; ++i) {
c = data[i];
if(c == ISO_gt) {
/* Full tag found. */
endtagfound();
parse_tag();
s.minorstate = MINORSTATE_TEXT;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
s.tagattrptr = 0;
endtagfound();
parse_tag();
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
s.tagptr = 0;
endtagfound();
break;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
} else if(iswhitespace(c) && s.tagattrparamptr == 0) {
/* Discard leading spaces. */
} else if((c == ISO_citation ||
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
c == ISO_citation2) && s.tagattrparamptr == 0) {
s.minorstate = MINORSTATE_TAGATTRPARAM;
s.quotechar = c;
PRINTF(("tag attr param q found\n"));
break;
} else if(iswhitespace(c)) {
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
PRINTF(("Non-leading space found at %d\n", s.tagattrparamptr));
/* Stop parsing if a non-leading space was found */
endtagfound();
parse_tag();
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
s.minorstate = MINORSTATE_TAGATTR;
s.tagattrptr = 0;
endtagfound();
break;
} else {
s.tagattrparam[s.tagattrparamptr] = c;
++s.tagattrparamptr;
/* Check if the "tagattr" field is full. If so, we just eat
up any data left in the tag. */
if(s.tagattrparamptr >= sizeof(s.tagattrparam) - 1) {
s.minorstate = MINORSTATE_TAGEND;
break;
}
}
}
break;
case MINORSTATE_TAGATTRPARAM:
/* We are parsing the "tag attr parameter", i.e., the link
part in <a href="link">. */
for(i = 0; i < len; ++i) {
c = data[i];
if(c == s.quotechar) {
/* Found end of tag attr parameter. */
endtagfound();
parse_tag();
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
s.minorstate = MINORSTATE_TAGATTR;
s.tagattrptr = 0;
endtagfound();
break;
} else {
if(iswhitespace(c)) {
s.tagattrparam[s.tagattrparamptr] = ISO_space;
} else {
s.tagattrparam[s.tagattrparamptr] = c;
}
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
++s.tagattrparamptr;
/* Check if the "tagattr" field is full. If so, we just eat
up any data left in the tag. */
if(s.tagattrparamptr >= sizeof(s.tagattrparam) - 1) {
s.minorstate = MINORSTATE_TAGEND;
break;
}
}
}
break;
case MINORSTATE_HTMLCOMMENT:
for(i = 0; i < len; ++i) {
c = data[i];
if(c == ISO_dash) {
++s.tagptr;
} else if(c == ISO_gt && s.tagptr > 0) {
PRINTF(("Comment done.\n"));
s.minorstate = MINORSTATE_TEXT;
break;
} else {
s.tagptr = 0;
}
}
break;
case MINORSTATE_TAGEND:
/* Discard characters until a '>' is seen. */
for(i = 0; i < len; ++i) {
if(data[i] == ISO_gt) {
s.minorstate = MINORSTATE_TEXT;
s.tagattrptr = 0;
endtagfound();
parse_tag();
break;
}
}
break;
default:
i = 0;
break;
}
if(i >= len) {
return len;
}
return i + 1;
}
/*-----------------------------------------------------------------------------------*/
void
htmlparser_parse(char *data, uint16_t datalen)
{
uint16_t plen;
while(datalen > 0) {
if(datalen > 255) {
plen = parse_word(data, 255);
} else {
plen = parse_word(data, (uint8_t)datalen);
}
datalen -= plen;
data += plen;
Reorganized web page attribute data handling. - Up to now the web browser used several fixed size arrays to hold the various types attribute data of the web page. This turned out to be way to inflexible for any non-trivial web page. Therefore now all attribute data is stored in a single buffer one after the other as they arrive from the parser only occupying the memory actually needed. This allows for pages with many links with rather short URLs as well as pages with few link with long URLs as well as pages with several simple forms as well as pages with one form with many form inputs. - Using the actual web page buffer to hold the text buffers of text entry fields was in general a cool idea but in reality it is often necessary to enter text longer than the size of the text entry field. Therefore the text buffer is now stored in the new unified attribute data buffer. - Splitting up the process of canonicalizing a link URL and actually navigating to the resulting URL allowed to get rid of the 'tmpurl' buffer used during form submit. Now the form action is canonicalized like a usual link, then the form input name/value pairs are written right into the 'url' buffer and afterwards the navigation is triggered. - Support for the 'render states' was completely removed. The only actually supported render state was centered output. The new unified attribute buffer would have complicated enumerating all widgets added to the page in order to adjust their position. Therefore I decided to drop the whole feature as the <center> tag is barely used anymore and newer center attributes are to hard to parse.
2013-03-06 15:29:36 +00:00
}
}
/*-----------------------------------------------------------------------------------*/