sax2.c.svn-base

这是一个用于解析xml文件的类库。使用这个类库

/*
 * SAX2.c : Default SAX2 handler to build a tree.
 *
 * See Copyright for the status of this software.
 *
 * Daniel Veillard <[e-mail]>
 */


#define IN_LIBXML
#include "libxml.h"
#include <stdlib.h>
#include <string.h>
#include <libxml/xmlmemory.h>
#include <libxml/tree.h>
#include <libxml/parser.h>
#include <libxml/parserInternals.h>
#include <libxml/valid.h>
#include <libxml/entities.h>
#include <libxml/xmlerror.h>
#include <libxml/debugXML.h>
#include <libxml/xmlIO.h>
#include <libxml/SAX.h>
#include <libxml/uri.h>
#include <libxml/valid.h>
#include <libxml/HTMLtree.h>
#include <libxml/globals.h>

/* #define DEBUG_SAX2 */
/* #define DEBUG_SAX2_TREE */

/**
 * TODO:
 *
 * macro to flag unimplemented blocks
 * XML_CATALOG_PREFER user env to select between system/public prefered
 * option. C.f. Richard Tobin <[e-mail]>
 *> Just FYI, I am using an environment variable XML_CATALOG_PREFER with
 *> values "system" and "public".  I have made the default be "system" to
 *> match yours.
 */
#define TODO 								\
    xmlGenericError(xmlGenericErrorContext,				\
	    "Unimplemented block at %s:%d\n",				\
            __FILE__, __LINE__);

/**
 * xmlValidError:
 * @ctxt:  an XML validation parser context
 * @error:  the error number
 * @msg:  the error message
 * @str1:  extra data
 * @str2:  extra data
 *
 * Handle a validation error
 */
static void
xmlErrValid(xmlParserCtxtPtr ctxt, xmlParserErrors error,
            const char *msg, const char *str1, const char *str2)
{
    xmlStructuredErrorFunc schannel = NULL;

    if ((ctxt != NULL) && (ctxt->disableSAX != 0) &&
        (ctxt->instate == XML_PARSER_EOF))
	return;
    ctxt->errNo = error;
    if ((ctxt->sax != NULL) && (ctxt->sax->initialized == XML_SAX2_MAGIC))
        schannel = ctxt->sax->serror;
    __xmlRaiseError(schannel,
                    ctxt->vctxt.error, ctxt->vctxt.userData,
                    ctxt, NULL, XML_FROM_DTD, error,
                    XML_ERR_ERROR, NULL, 0, (const char *) str1,
		    (const char *) str2, NULL, 0, 0,
		    msg, (const char *) str1, (const char *) str2);
    ctxt->valid = 0;
}

/**
 * xmlSAX2GetPublicId:
 * @ctx: the user data (XML parser context)
 *
 * Provides the public ID e.g. "-//SGMLSOURCE//DTD DEMO//EN"
 *
 * Returns a xmlChar *
 */
const xmlChar *
xmlSAX2GetPublicId(void *ctx ATTRIBUTE_UNUSED)
{
    /* xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx; */
    return(NULL);
}

/**
 * xmlSAX2GetSystemId:
 * @ctx: the user data (XML parser context)
 *
 * Provides the system ID, basically URL or filename e.g.
 * http://www.sgmlsource.com/dtds/memo.dtd
 *
 * Returns a xmlChar *
 */
const xmlChar *
xmlSAX2GetSystemId(void *ctx)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    return((const xmlChar *) ctxt->input->filename); 
}

/**
 * xmlSAX2GetLineNumber:
 * @ctx: the user data (XML parser context)
 *
 * Provide the line number of the current parsing point.
 *
 * Returns an int
 */
int
xmlSAX2GetLineNumber(void *ctx)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    return(ctxt->input->line);
}

/**
 * xmlSAX2GetColumnNumber:
 * @ctx: the user data (XML parser context)
 *
 * Provide the column number of the current parsing point.
 *
 * Returns an int
 */
int
xmlSAX2GetColumnNumber(void *ctx)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    return(ctxt->input->col);
}

/**
 * xmlSAX2IsStandalone:
 * @ctx: the user data (XML parser context)
 *
 * Is this document tagged standalone ?
 *
 * Returns 1 if true
 */
int
xmlSAX2IsStandalone(void *ctx)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    return(ctxt->myDoc->standalone == 1);
}

/**
 * xmlSAX2HasInternalSubset:
 * @ctx: the user data (XML parser context)
 *
 * Does this document has an internal subset
 *
 * Returns 1 if true
 */
int
xmlSAX2HasInternalSubset(void *ctx)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    return(ctxt->myDoc->intSubset != NULL);
}

/**
 * xmlSAX2HasExternalSubset:
 * @ctx: the user data (XML parser context)
 *
 * Does this document has an external subset
 *
 * Returns 1 if true
 */
int
xmlSAX2HasExternalSubset(void *ctx)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    return(ctxt->myDoc->extSubset != NULL);
}

/**
 * xmlSAX2InternalSubset:
 * @ctx:  the user data (XML parser context)
 * @name:  the root element name
 * @ExternalID:  the external ID
 * @SystemID:  the SYSTEM ID (e.g. filename or URL)
 *
 * Callback on internal subset declaration.
 */
void
xmlSAX2InternalSubset(void *ctx, const xmlChar *name,
	       const xmlChar *ExternalID, const xmlChar *SystemID)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    xmlDtdPtr dtd;
#ifdef DEBUG_SAX
    xmlGenericError(xmlGenericErrorContext,
	    "SAX.xmlSAX2InternalSubset(%s, %s, %s)\n",
            name, ExternalID, SystemID);
#endif

    if (ctxt->myDoc == NULL)
	return;
    dtd = xmlGetIntSubset(ctxt->myDoc);
    if (dtd != NULL) {
	if (ctxt->html)
	    return;
	xmlUnlinkNode((xmlNodePtr) dtd);
	xmlFreeDtd(dtd);
	ctxt->myDoc->intSubset = NULL;
    }
    ctxt->myDoc->intSubset = 
	xmlCreateIntSubset(ctxt->myDoc, name, ExternalID, SystemID);
}

/**
 * xmlSAX2ExternalSubset:
 * @ctx: the user data (XML parser context)
 * @name:  the root element name
 * @ExternalID:  the external ID
 * @SystemID:  the SYSTEM ID (e.g. filename or URL)
 *
 * Callback on external subset declaration.
 */
void
xmlSAX2ExternalSubset(void *ctx, const xmlChar *name,
	       const xmlChar *ExternalID, const xmlChar *SystemID)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
#ifdef DEBUG_SAX
    xmlGenericError(xmlGenericErrorContext,
	    "SAX.xmlSAX2ExternalSubset(%s, %s, %s)\n",
            name, ExternalID, SystemID);
#endif
    if (((ExternalID != NULL) || (SystemID != NULL)) &&
        (((ctxt->validate) || (ctxt->loadsubset != 0)) &&
	 (ctxt->wellFormed && ctxt->myDoc))) {
	/*
	 * Try to fetch and parse the external subset.
	 */
	xmlParserInputPtr oldinput;
	int oldinputNr;
	int oldinputMax;
	xmlParserInputPtr *oldinputTab;
	xmlParserInputPtr input = NULL;
	xmlCharEncoding enc;
	int oldcharset;

	/*
	 * Ask the Entity resolver to load the damn thing
	 */
	if ((ctxt->sax != NULL) && (ctxt->sax->resolveEntity != NULL))
	    input = ctxt->sax->resolveEntity(ctxt->userData, ExternalID,
	                                        SystemID);
	if (input == NULL) {
	    return;
	}

	xmlNewDtd(ctxt->myDoc, name, ExternalID, SystemID);

	/*
	 * make sure we won't destroy the main document context
	 */
	oldinput = ctxt->input;
	oldinputNr = ctxt->inputNr;
	oldinputMax = ctxt->inputMax;
	oldinputTab = ctxt->inputTab;
	oldcharset = ctxt->charset;

	ctxt->inputTab = (xmlParserInputPtr *)
	                 xmlMalloc(5 * sizeof(xmlParserInputPtr));
	if (ctxt->inputTab == NULL) {
	    ctxt->errNo = XML_ERR_NO_MEMORY;
	    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		ctxt->sax->error(ctxt->userData, 
		     "xmlSAX2ExternalSubset: out of memory\n");
	    ctxt->errNo = XML_ERR_NO_MEMORY;
	    ctxt->instate = XML_PARSER_EOF;
	    ctxt->disableSAX = 1;
	    ctxt->input = oldinput;
	    ctxt->inputNr = oldinputNr;
	    ctxt->inputMax = oldinputMax;
	    ctxt->inputTab = oldinputTab;
	    ctxt->charset = oldcharset;
	    return;
	}
	ctxt->inputNr = 0;
	ctxt->inputMax = 5;
	ctxt->input = NULL;
	xmlPushInput(ctxt, input);

	/*
	 * On the fly encoding conversion if needed
	 */
	if (ctxt->input->length >= 4) {
	    enc = xmlDetectCharEncoding(ctxt->input->cur, 4);
	    xmlSwitchEncoding(ctxt, enc);
	}

	if (input->filename == NULL)
	    input->filename = (char *) xmlCanonicPath(SystemID);
	input->line = 1;
	input->col = 1;
	input->base = ctxt->input->cur;
	input->cur = ctxt->input->cur;
	input->free = NULL;

	/*
	 * let's parse that entity knowing it's an external subset.
	 */
	xmlParseExternalSubset(ctxt, ExternalID, SystemID);

        /*
	 * Free up the external entities
	 */

	while (ctxt->inputNr > 1)
	    xmlPopInput(ctxt);
	xmlFreeInputStream(ctxt->input);
        xmlFree(ctxt->inputTab);

	/*
	 * Restore the parsing context of the main entity
	 */
	ctxt->input = oldinput;
	ctxt->inputNr = oldinputNr;
	ctxt->inputMax = oldinputMax;
	ctxt->inputTab = oldinputTab;
	ctxt->charset = oldcharset;
	/* ctxt->wellFormed = oldwellFormed; */
    }
}

/**
 * xmlSAX2ResolveEntity:
 * @ctx: the user data (XML parser context)
 * @publicId: The public ID of the entity
 * @systemId: The system ID of the entity
 *
 * The entity loader, to control the loading of external entities,
 * the application can either:
 *    - override this xmlSAX2ResolveEntity() callback in the SAX block
 *    - or better use the xmlSetExternalEntityLoader() function to
 *      set up it's own entity resolution routine
 *
 * Returns the xmlParserInputPtr if inlined or NULL for DOM behaviour.
 */
xmlParserInputPtr
xmlSAX2ResolveEntity(void *ctx, const xmlChar *publicId, const xmlChar *systemId)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    xmlParserInputPtr ret;
    xmlChar *URI;
    const char *base = NULL;

    if (ctxt->input != NULL)
	base = ctxt->input->filename;
    if (base == NULL)
	base = ctxt->directory;

    URI = xmlBuildURI(systemId, (const xmlChar *) base);

#ifdef DEBUG_SAX
    xmlGenericError(xmlGenericErrorContext,
	    "SAX.xmlSAX2ResolveEntity(%s, %s)\n", publicId, systemId);
#endif

    ret = xmlLoadExternalEntity((const char *) URI,
				(const char *) publicId, ctxt);
    if (URI != NULL)
	xmlFree(URI);
    return(ret);
}

/**
 * xmlSAX2GetEntity:
 * @ctx: the user data (XML parser context)
 * @name: The entity name
 *
 * Get an entity by name
 *
 * Returns the xmlEntityPtr if found.
 */
xmlEntityPtr
xmlSAX2GetEntity(void *ctx, const xmlChar *name)
{
    xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
    xmlEntityPtr ret = NULL;

#ifdef DEBUG_SAX
    xmlGenericError(xmlGenericErrorContext,
	    "SAX.xmlSAX2GetEntity(%s)\n", name);
#endif

    if (ctxt->inSubset == 0) {
	ret = xmlGetPredefinedEntity(name);
	if (ret != NULL)
	    return(ret);
    }
    if ((ctxt->myDoc != NULL) && (ctxt->myDoc->standalone == 1)) {
	if (ctxt->inSubset == 2) {
	    ctxt->myDoc->standalone = 0;
	    ret = xmlGetDocEntity(ctxt->myDoc, name);
	    ctxt->myDoc->standalone = 1;
	} else {
	    ret = xmlGetDocEntity(ctxt->myDoc, name);
	    if (ret == NULL) {
		ctxt->myDoc->standalone = 0;
		ret = xmlGetDocEntity(ctxt->myDoc, name);
		if (ret != NULL) {
		    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
			ctxt->sax->error(ctxt->userData, 
		 "Entity(%s) document marked standalone but requires external subset\n",
					 name);
		    ctxt->valid = 0;
		    ctxt->wellFormed = 0;
		}
		ctxt->myDoc->standalone = 1;
	    }
	}
    } else {
	ret = xmlGetDocEntity(ctxt->myDoc, name);
    }
    if ((ret != NULL) &&
	((ctxt->validate) || (ctxt->replaceEntities)) &&
	(ret->children == NULL) &&
	(ret->etype == XML_EXTERNAL_GENERAL_PARSED_ENTITY)) {
	int val;

	/*
	 * for validation purposes we really need to fetch and
	 * parse the external entity
	 */
	xmlNodePtr children;

        val = xmlParseCtxtExternalEntity(ctxt, ret->URI,
		                         ret->ExternalID, &children);
	if (val == 0) {
	    xmlAddChildList((xmlNodePtr) ret, children);
	} else {
	    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	        ctxt->sax->error(ctxt->userData, 
		 "Failure to process entity %s\n", name);
	    ctxt->wellFormed = 0;
	    ctxt->valid = 0;
	    ctxt->validate = 0;
	    return(NULL);
	}
	ret->owner = 1;
    }
    return(ret);
}

/**
 * xmlSAX2GetParameterEntity:
 * @ctx: the user data (XML parser context)
 * @name: The entity name
 *
 * Get a parameter entity by name
 *
 * Returns the xmlEntityPtr if found.