htmlparser.c

SIP（Session Initiation Protocol）是由IETF定义

/*
 * HTMLparser.c : an HTML 4.0 non-verifying parser
 *
 * See Copyright for the status of this software.
 *
 * [e-mail]
 */

#ifdef WIN32
#include "win32config.h"
#else
#include "config.h"
#endif

#include "xmlversion.h"
#ifdef LIBXML_HTML_ENABLED

#include <stdio.h>
#include <string.h> /* for memset() only */
#ifdef HAVE_CTYPE_H
#include <ctype.h>
#endif
#ifdef HAVE_STDLIB_H
#include <stdlib.h>
#endif
#ifdef HAVE_SYS_STAT_H
#include <sys/stat.h>
#endif
#ifdef HAVE_FCNTL_H
#include <fcntl.h>
#endif
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif
#ifdef HAVE_ZLIB_H
#include <zlib.h>
#endif

#include <libxml/xmlmemory.h>
#include <libxml/tree.h>
#include <libxml/HTMLparser.h>
#include <libxml/entities.h>
#include <libxml/encoding.h>
#include <libxml/valid.h>
#include <libxml/parserInternals.h>
#include <libxml/xmlIO.h>
#include "xml-error.h"

#define HTML_MAX_NAMELEN 1000
#define INPUT_CHUNK     50
#define HTML_PARSER_BIG_BUFFER_SIZE 1024
#define HTML_PARSER_BUFFER_SIZE 100

/* #define DEBUG */
/* #define DEBUG_PUSH */

/************************************************************************
 *									*
 * 		Parser stacks related functions and macros		*
 *									*
 ************************************************************************/

/*
 * Generic function for accessing stacks in the Parser Context
 */

#define PUSH_AND_POP(scope, type, name)					\
scope int html##name##Push(htmlParserCtxtPtr ctxt, type value) {	\
    if (ctxt->name##Nr >= ctxt->name##Max) {				\
	ctxt->name##Max *= 2;						\
        ctxt->name##Tab = (void *) xmlRealloc(ctxt->name##Tab,		\
	             ctxt->name##Max * sizeof(ctxt->name##Tab[0]));	\
        if (ctxt->name##Tab == NULL) {					\
	    fprintf(stderr, "realloc failed !\n");			\
	    return(0);							\
	}								\
    }									\
    ctxt->name##Tab[ctxt->name##Nr] = value;				\
    ctxt->name = value;							\
    return(ctxt->name##Nr++);						\
}									\
scope type html##name##Pop(htmlParserCtxtPtr ctxt) {			\
    type ret;								\
    if (ctxt->name##Nr < 0) return(0);					\
    ctxt->name##Nr--;							\
    if (ctxt->name##Nr < 0) return(0);					\
    if (ctxt->name##Nr > 0)						\
	ctxt->name = ctxt->name##Tab[ctxt->name##Nr - 1];		\
    else								\
        ctxt->name = NULL;						\
    ret = ctxt->name##Tab[ctxt->name##Nr];				\
    ctxt->name##Tab[ctxt->name##Nr] = 0;				\
    return(ret);							\
}									\

PUSH_AND_POP(extern, xmlNodePtr, node)
PUSH_AND_POP(extern, xmlChar*, name)

/*
 * Macros for accessing the content. Those should be used only by the parser,
 * and not exported.
 *
 * Dirty macros, i.e. one need to make assumption on the context to use them
 *
 *   CUR_PTR return the current pointer to the xmlChar to be parsed.
 *   CUR     returns the current xmlChar value, i.e. a 8 bit value if compiled
 *           in ISO-Latin or UTF-8, and the current 16 bit value if compiled
 *           in UNICODE mode. This should be used internally by the parser
 *           only to compare to ASCII values otherwise it would break when
 *           running with UTF-8 encoding.
 *   NXT(n)  returns the n'th next xmlChar. Same as CUR is should be used only
 *           to compare on ASCII based substring.
 *   UPP(n)  returns the n'th next xmlChar converted to uppercase. Same as CUR
 *           it should be used only to compare on ASCII based substring.
 *   SKIP(n) Skip n xmlChar, and must also be used only to skip ASCII defined
 *           strings within the parser.
 *
 * Clean macros, not dependent of an ASCII context, expect UTF-8 encoding
 *
 *   CURRENT Returns the current char value, with the full decoding of
 *           UTF-8 if we are using this mode. It returns an int.
 *   NEXT    Skip to the next character, this does the proper decoding
 *           in UTF-8 mode. It also pop-up unfinished entities on the fly.
 *   COPY(to) copy one char to *to, increment CUR_PTR and to accordingly
 */

#define CUR ((int) (*ctxt->input->cur))
    
#define UPPER (toupper(*ctxt->input->cur))

#define SKIP(val) ctxt->nbChars += (val),ctxt->input->cur += (val)

#define NXT(val) ctxt->input->cur[(val)]

#define UPP(val) (toupper(ctxt->input->cur[(val)]))

#define CUR_PTR ctxt->input->cur

#define SHRINK  xmlParserInputShrink(ctxt->input)

#define GROW  xmlParserInputGrow(ctxt->input, INPUT_CHUNK)

#define CURRENT ((int) (*ctxt->input->cur))

#define NEXT htmlNextChar(ctxt);

#define SKIP_BLANKS htmlSkipBlankChars(ctxt);

/**
 * htmlNextChar:
 * @ctxt:  the HTML parser context
 *
 * Skip to the next char input char.
 */

void
htmlNextChar(htmlParserCtxtPtr ctxt) {
    if ((*ctxt->input->cur == 0) &&
        (xmlParserInputGrow(ctxt->input, INPUT_CHUNK) <= 0)) {
	    xmlPopInput(ctxt);
    } else {
        if (*(ctxt->input->cur) == '\n') {
	    ctxt->input->line++; ctxt->input->col = 1;
	} else ctxt->input->col++;
	ctxt->input->cur++;
	ctxt->nbChars++;
        if (*ctxt->input->cur == 0)
	    xmlParserInputGrow(ctxt->input, INPUT_CHUNK);
    }
}

/**
 * htmlSkipBlankChars:
 * @ctxt:  the HTML parser context
 *
 * skip all blanks character found at that point in the input streams.
 *
 * Returns the number of space chars skipped
 */

int
htmlSkipBlankChars(xmlParserCtxtPtr ctxt) {
    int res = 0;

    while (IS_BLANK(*(ctxt->input->cur))) {
	if ((*ctxt->input->cur == 0) &&
	    (xmlParserInputGrow(ctxt->input, INPUT_CHUNK) <= 0)) {
		xmlPopInput(ctxt);
	} else {
	    if (*(ctxt->input->cur) == '\n') {
		ctxt->input->line++; ctxt->input->col = 1;
	    } else ctxt->input->col++;
	    ctxt->input->cur++;
	    ctxt->nbChars++;
	    if (*ctxt->input->cur == 0)
		xmlParserInputGrow(ctxt->input, INPUT_CHUNK);
	}
	res++;
    }
    return(res);
}



/************************************************************************
 *									*
 * 		The list of HTML elements and their properties		*
 *									*
 ************************************************************************/

/*
 *  Start Tag: 1 means the start tag can be ommited
 *  End Tag:   1 means the end tag can be ommited
 *             2 means it's forbidden (empty elements)
 *  Depr:      this element is deprecated
 *  DTD:       1 means that this element is valid only in the Loose DTD
 *             2 means that this element is valid only in the Frameset DTD
 *
 * Name,Start Tag,End Tag,  Empty,  Depr.,    DTD, Description
 */
htmlElemDesc  html40ElementTable[] = {
{ "a",		0,	0,	0,	0,	0, "anchor " },
{ "abbr",	0,	0,	0,	0,	0, "abbreviated form" },
{ "acronym",	0,	0,	0,	0,	0, "" },
{ "address",	0,	0,	0,	0,	0, "information on author " },
{ "applet",	0,	0,	0,	1,	1, "java applet " },
{ "area",	0,	2,	1,	0,	0, "client-side image map area " },
{ "b",		0,	0,	0,	0,	0, "bold text style" },
{ "base",	0,	2,	1,	0,	0, "document base uri " },
{ "basefont",	0,	2,	1,	1,	1, "base font size " },
{ "bdo",	0,	0,	0,	0,	0, "i18n bidi over-ride " },
{ "big",	0,	0,	0,	0,	0, "large text style" },
{ "blockquote",	0,	0,	0,	0,	0, "long quotation " },
{ "body",	1,	1,	0,	0,	0, "document body " },
{ "br",		0,	2,	1,	0,	0, "forced line break " },
{ "button",	0,	0,	0,	0,	0, "push button " },
{ "caption",	0,	0,	0,	0,	0, "table caption " },
{ "center",	0,	0,	0,	1,	1, "shorthand for div align=center " },
{ "cite",	0,	0,	0,	0,	0, "citation" },
{ "code",	0,	0,	0,	0,	0, "computer code fragment" },
{ "col",	0,	2,	1,	0,	0, "table column " },
{ "colgroup",	0,	1,	0,	0,	0, "table column group " },
{ "dd",		0,	1,	0,	0,	0, "definition description " },
{ "del",	0,	0,	0,	0,	0, "deleted text " },
{ "dfn",	0,	0,	0,	0,	0, "instance definition" },
{ "dir",	0,	0,	0,	1,	1, "directory list" },
{ "div",	0,	0,	0,	0,	0, "generic language/style container"},
{ "dl",		0,	0,	0,	0,	0, "definition list " },
{ "dt",		0,	1,	0,	0,	0, "definition term " },
{ "em",		0,	0,	0,	0,	0, "emphasis" },
{ "fieldset",	0,	0,	0,	0,	0, "form control group " },
{ "font",	0,	0,	0,	1,	1, "local change to font " },
{ "form",	0,	0,	0,	0,	0, "interactive form " },
{ "frame",	0,	2,	1,	0,	2, "subwindow " },
{ "frameset",	0,	0,	0,	0,	2, "window subdivision" },
{ "h1",		0,	0,	0,	0,	0, "heading " },
{ "h2",		0,	0,	0,	0,	0, "heading " },
{ "h3",		0,	0,	0,	0,	0, "heading " },
{ "h4",		0,	0,	0,	0,	0, "heading " },
{ "h5",		0,	0,	0,	0,	0, "heading " },
{ "h6",		0,	0,	0,	0,	0, "heading " },
{ "head",	1,	1,	0,	0,	0, "document head " },
{ "hr",		0,	2,	1,	0,	0, "horizontal rule " },
{ "html",	1,	1,	0,	0,	0, "document root element " },
{ "i",		0,	0,	0,	0,	0, "italic text style" },
{ "iframe",	0,	0,	0,	0,	1, "inline subwindow " },
{ "img",	0,	2,	1,	0,	0, "embedded image " },
{ "input",	0,	2,	1,	0,	0, "form control " },
{ "ins",	0,	0,	0,	0,	0, "inserted text" },
{ "isindex",	0,	2,	1,	1,	1, "single line prompt " },
{ "kbd",	0,	0,	0,	0,	0, "text to be entered by the user" },
{ "label",	0,	0,	0,	0,	0, "form field label text " },
{ "legend",	0,	0,	0,	0,	0, "fieldset legend " },
{ "li",		0,	1,	0,	0,	0, "list item " },
{ "link",	0,	2,	1,	0,	0, "a media-independent link " },
{ "map",	0,	0,	0,	0,	0, "client-side image map " },
{ "menu",	0,	0,	0,	1,	1, "menu list " },
{ "meta",	0,	2,	1,	0,	0, "generic metainformation " },
{ "noframes",	0,	0,	0,	0,	2, "alternate content container for non frame-based rendering " },
{ "noscript",	0,	0,	0,	0,	0, "alternate content container for non script-based rendering " },
{ "object",	0,	0,	0,	0,	0, "generic embedded object " },
{ "ol",		0,	0,	0,	0,	0, "ordered list " },
{ "optgroup",	0,	0,	0,	0,	0, "option group " },
{ "option",	0,	1,	0,	0,	0, "selectable choice " },
{ "p",		0,	1,	0,	0,	0, "paragraph " },
{ "param",	0,	2,	1,	0,	0, "named property value " },
{ "pre",	0,	0,	0,	0,	0, "preformatted text " },
{ "q",		0,	0,	0,	0,	0, "short inline quotation " },
{ "s",		0,	0,	0,	1,	1, "strike-through text style" },
{ "samp",	0,	0,	0,	0,	0, "sample program output, scripts, etc." },
{ "script",	0,	0,	0,	0,	0, "script statements " },
{ "select",	0,	0,	0,	0,	0, "option selector " },
{ "small",	0,	0,	0,	0,	0, "small text style" },
{ "span",	0,	0,	0,	0,	0, "generic language/style container " },
{ "strike",	0,	0,	0,	1,	1, "strike-through text" },
{ "strong",	0,	0,	0,	0,	0, "strong emphasis" },
{ "style",	0,	0,	0,	0,	0, "style info " },
{ "sub",	0,	0,	0,	0,	0, "subscript" },
{ "sup",	0,	0,	0,	0,	0, "superscript " },
{ "table",	0,	0,	0,	0,	0, "&#160;" },
{ "tbody",	1,	1,	0,	0,	0, "table body " },
{ "td",		0,	1,	0,	0,	0, "table data cell" },
{ "textarea",	0,	0,	0,	0,	0, "multi-line text field " },
{ "tfoot",	0,	1,	0,	0,	0, "table footer " },
{ "th",		0,	1,	0,	0,	0, "table header cell" },
{ "thead",	0,	1,	0,	0,	0, "table header " },
{ "title",	0,	0,	0,	0,	0, "document title " },
{ "tr",		0,	1,	0,	0,	0, "table row " },
{ "tt",		0,	0,	0,	0,	0, "teletype or monospaced text style" },
{ "u",		0,	0,	0,	1,	1, "underlined text style" },
{ "ul",		0,	0,	0,	0,	0, "unordered list " },
{ "var",	0,	0,	0,	0,	0, "instance of a variable or program argument" },
};

/*
 * start tags that imply the end of a current element
 * any tag of each line implies the end of the current element if the type of
 * that element is in the same line
 */
char *htmlEquEnd[] = {
"dt", "dd", "li", "option", NULL,
"h1", "h2", "h3", "h4", "h5", "h6", NULL,
"ol", "menu", "dir", "address", "pre", "listing", "xmp", NULL,
NULL
};
/*
 * acording the HTML DTD, HR should be added to the 2nd line above, as it
 * is not allowed within a H1, H2, H3, etc. But we should tolerate that case
 * because many documents contain rules in headings...
 */

/*
 * start tags that imply the end of current element
 */
char *htmlStartClose[] = {
"form",		"form", "p", "hr", "h1", "h2", "h3", "h4", "h5", "h6",
		"dl", "ul", "ol", "menu", "dir", "address", "pre",
		"listing", "xmp", "head", NULL,
"head",		"p", NULL,
"title",	"p", NULL,
"body",		"head", "style", "link", "title", "p", NULL,
"li",		"p", "h1", "h2", "h3", "h4", "h5", "h6", "dl", "address",
		"pre", "listing", "xmp", "head", "li", NULL,
"hr",		"p", "head", NULL,
"h1",		"p", "head", NULL,
"h2",		"p", "head", NULL,
"h3",		"p", "head", NULL,
"h4",		"p", "head", NULL,
"h5",		"p", "head", NULL,
"h6",		"p", "head", NULL,
"dir",		"p", "head", NULL,
"address",	"p", "head", "ul", NULL,
"pre",		"p", "head", "ul", NULL,
"listing",	"p", "head", NULL,
"xmp",		"p", "head", NULL,
"blockquote",	"p", "head", NULL,
"dl",		"p", "dt", "menu", "dir", "address", "pre", "listing",
		"xmp", "head", NULL,
"dt",		"p", "menu", "dir", "address", "pre", "listing", "xmp",
                "head", "dd", NULL,
"dd",		"p", "menu", "dir", "address", "pre", "listing", "xmp",
                "head", "dt", NULL,
"ul",		"p", "head", "ol", "menu", "dir", "address", "pre",
		"listing", "xmp", NULL,
"ol",		"p", "head", "ul", NULL,
"menu",		"p", "head", "ul", NULL,
"p",		"p", "head", "h1", "h2", "h3", "h4", "h5", "h6", NULL,
"div",		"p", "head", NULL,
"noscript",	"p", "head", NULL,
"center",	"font", "b", "i", "p", "head", NULL,
"a",		"a", NULL,
"caption",	"p", NULL,
"colgroup",	"caption", "colgroup", "col", "p", NULL,
"col",		"caption", "col", "p", NULL,
"table",	"p", "head", "h1", "h2", "h3", "h4", "h5", "h6", "pre",
		"listing", "xmp", "a", NULL,
"th",		"th", "td", NULL,
"td",		"th", "td", "p", NULL,
"tr",		"th", "td", "tr", "caption", "col", "colgroup", "p", NULL,
"thead",	"caption", "col", "colgroup", NULL,
"tfoot",	"th", "td", "tr", "caption", "col", "colgroup", "thead",
		"tbody", "p", NULL,
"tbody",	"th", "td", "tr", "caption", "col", "colgroup", "thead",
		"tfoot", "tbody", "p", NULL,
"optgroup",	"option", NULL,
"fieldset",	"legend", "p", "head", "h1", "h2", "h3", "h4", "h5", "h6",
		"pre", "listing", "xmp", "a", NULL,
NULL
};

static char** htmlStartCloseIndex[100];
static int htmlStartCloseIndexinitialized = 0;

/************************************************************************
 *									*
 * 		functions to handle HTML specific data			*
 *									*
 ************************************************************************/

/**