parserinternals.h.svn-base

这是一个用于解析xml文件的类库。使用这个类库

/*
 * Summary: internals routines exported by the parser.
 * Description: this module exports a number of internal parsing routines
 *              they are not really all intended for applications but
 *              can prove useful doing low level processing.
 *
 * Copy: See Copyright for the status of this software.
 *
 * Author: Daniel Veillard
 */

#ifndef __XML_PARSER_INTERNALS_H__
#define __XML_PARSER_INTERNALS_H__

#include <libxml/xmlversion.h>
#include <libxml/parser.h>
#include <libxml/HTMLparser.h>
#include <libxml/chvalid.h>

#ifdef __cplusplus
extern "C" {
#endif

/**
 * xmlParserMaxDepth:
 *
 * arbitrary depth limit for the XML documents that we allow to 
 * process. This is not a limitation of the parser but a safety 
 * boundary feature.
 */
XMLPUBVAR unsigned int xmlParserMaxDepth;

 /**
  * XML_MAX_NAMELEN:
  *
  * Identifiers can be longer, but this will be more costly
  * at runtime.
  */
#define XML_MAX_NAMELEN 100

/**
 * INPUT_CHUNK:
 *
 * The parser tries to always have that amount of input ready.
 * One of the point is providing context when reporting errors.
 */
#define INPUT_CHUNK	250

/************************************************************************
 *									*
 * UNICODE version of the macros.      					*
 *									*
 ************************************************************************/
/**
 * IS_BYTE_CHAR:
 * @c:  an byte value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 * [2] Char ::= #x9 | #xA | #xD | [#x20...]
 * any byte character in the accepted range
 */
#define IS_BYTE_CHAR(c)	 xmlIsChar_ch(c)

/**
 * IS_CHAR:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 * [2] Char ::= #x9 | #xA | #xD | [#x20-#xD7FF] | [#xE000-#xFFFD]
 *                  | [#x10000-#x10FFFF]
 * any Unicode character, excluding the surrogate blocks, FFFE, and FFFF.
 */
#define IS_CHAR(c)   xmlIsCharQ(c)

/**
 * IS_CHAR_CH:
 * @c: an xmlChar (usually an unsigned char)
 *
 * Behaves like IS_CHAR on single-byte value
 */
#define IS_CHAR_CH(c)  xmlIsChar_ch(c)

/**
 * IS_BLANK:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 * [3] S ::= (#x20 | #x9 | #xD | #xA)+
 */
#define IS_BLANK(c)  xmlIsBlankQ(c)

/**
 * IS_BLANK_CH:
 * @c:  an xmlChar value (normally unsigned char)
 *
 * Behaviour same as IS_BLANK
 */
#define IS_BLANK_CH(c)  xmlIsBlank_ch(c)

/**
 * IS_BASECHAR:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 * [85] BaseChar ::= ... long list see REC ...
 */
#define IS_BASECHAR(c) xmlIsBaseCharQ(c)

/**
 * IS_DIGIT:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 * [88] Digit ::= ... long list see REC ...
 */
#define IS_DIGIT(c) xmlIsDigitQ(c)

/**
 * IS_DIGIT_CH:
 * @c:  an xmlChar value (usually an unsigned char)
 *
 * Behaves like IS_DIGIT but with a single byte argument
 */
#define IS_DIGIT_CH(c)  xmlIsDigit_ch(c)

/**
 * IS_COMBINING:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 * [87] CombiningChar ::= ... long list see REC ...
 */
#define IS_COMBINING(c) xmlIsCombiningQ(c)

/**
 * IS_COMBINING_CH:
 * @c:  an xmlChar (usually an unsigned char)
 *
 * Always false (all combining chars > 0xff)
 */
#define IS_COMBINING_CH(c) 0 

/**
 * IS_EXTENDER:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 *
 * [89] Extender ::= #x00B7 | #x02D0 | #x02D1 | #x0387 | #x0640 |
 *                   #x0E46 | #x0EC6 | #x3005 | [#x3031-#x3035] |
 *                   [#x309D-#x309E] | [#x30FC-#x30FE]
 */
#define IS_EXTENDER(c) xmlIsExtenderQ(c)

/**
 * IS_EXTENDER_CH:
 * @c:  an xmlChar value (usually an unsigned char)
 *
 * Behaves like IS_EXTENDER but with a single-byte argument
 */
#define IS_EXTENDER_CH(c)  xmlIsExtender_ch(c)

/**
 * IS_IDEOGRAPHIC:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 *
 * [86] Ideographic ::= [#x4E00-#x9FA5] | #x3007 | [#x3021-#x3029]
 */
#define IS_IDEOGRAPHIC(c) xmlIsIdeographicQ(c)

/**
 * IS_LETTER:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 *
 * [84] Letter ::= BaseChar | Ideographic 
 */
#define IS_LETTER(c) (IS_BASECHAR(c) || IS_IDEOGRAPHIC(c))

/**
 * IS_LETTER_CH:
 * @c:  an xmlChar value (normally unsigned char)
 *
 * Macro behaves like IS_LETTER, but only check base chars
 *
 */
#define IS_LETTER_CH(c) xmlIsBaseChar_ch(c)
/**
 * IS_PUBIDCHAR:
 * @c:  an UNICODE value (int)
 *
 * Macro to check the following production in the XML spec:
 *
 *
 * [13] PubidChar ::= #x20 | #xD | #xA | [a-zA-Z0-9] | [-'()+,./:=?;!*#@$_%]
 */
#define IS_PUBIDCHAR(c)	xmlIsPubidCharQ(c)

/**
 * IS_PUBIDCHAR_CH:
 * @c:  an xmlChar value (normally unsigned char)
 *
 * Same as IS_PUBIDCHAR but for single-byte value
 */
#define IS_PUBIDCHAR_CH(c) xmlIsPubidChar_ch(c)

/**
 * SKIP_EOL:
 * @p:  and UTF8 string pointer
 *
 * Skips the end of line chars.
 */
#define SKIP_EOL(p) 							\
    if (*(p) == 0x13) { p++ ; if (*(p) == 0x10) p++; }			\
    if (*(p) == 0x10) { p++ ; if (*(p) == 0x13) p++; }

/**
 * MOVETO_ENDTAG:
 * @p:  and UTF8 string pointer
 *
 * Skips to the next '>' char.
 */
#define MOVETO_ENDTAG(p)						\
    while ((*p) && (*(p) != '>')) (p)++

/**
 * MOVETO_STARTTAG:
 * @p:  and UTF8 string pointer
 *
 * Skips to the next '<' char.
 */
#define MOVETO_STARTTAG(p)						\
    while ((*p) && (*(p) != '<')) (p)++

/**
 * Global variables used for predefined strings.
 */
XMLPUBVAR const xmlChar xmlStringText[];
XMLPUBVAR const xmlChar xmlStringTextNoenc[];
XMLPUBVAR const xmlChar xmlStringComment[];

/*
 * Function to finish the work of the macros where needed.
 */
XMLPUBFUN int XMLCALL                   xmlIsLetter     (int c);

/**
 * Parser context.
 */
XMLPUBFUN xmlParserCtxtPtr XMLCALL	
			xmlCreateFileParserCtxt	(const char *filename);
XMLPUBFUN xmlParserCtxtPtr XMLCALL	
			xmlCreateURLParserCtxt	(const char *filename,
						 int options);
XMLPUBFUN xmlParserCtxtPtr XMLCALL	
			xmlCreateMemoryParserCtxt(const char *buffer,
						 int size);
XMLPUBFUN xmlParserCtxtPtr XMLCALL	
			xmlCreateEntityParserCtxt(const xmlChar *URL,
						 const xmlChar *ID,
						 const xmlChar *base);
XMLPUBFUN int XMLCALL			
			xmlSwitchEncoding	(xmlParserCtxtPtr ctxt,
						 xmlCharEncoding enc);
XMLPUBFUN int XMLCALL			
			xmlSwitchToEncoding	(xmlParserCtxtPtr ctxt,
					 xmlCharEncodingHandlerPtr handler);
XMLPUBFUN int XMLCALL			
			xmlSwitchInputEncoding	(xmlParserCtxtPtr ctxt,
						 xmlParserInputPtr input,
					 xmlCharEncodingHandlerPtr handler);

#ifdef IN_LIBXML
/* internal error reporting */
XMLPUBFUN void XMLCALL
			__xmlErrEncoding	(xmlParserCtxtPtr ctxt,
						 xmlParserErrors error,