parser.c

SIP 1.5.0源代码

	}
    } else if  ((RAW == '&') && (NXT(1) == '#')) {
	SKIP(2);
	while (RAW != ';') {
	    if ((RAW >= '0') && (RAW <= '9')) 
	        val = val * 10 + (CUR - '0');
	    else {
		ctxt->errNo = XML_ERR_INVALID_DEC_CHARREF;
	        if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		    ctxt->sax->error(ctxt->userData, 
		         "xmlParseCharRef: invalid decimal value\n");
		ctxt->wellFormed = 0;
		ctxt->disableSAX = 1;
		val = 0;
		break;
	    }
	    NEXT;
	}
	if (RAW == ';') {
	    /* on purpose to avoid reentrancy problems with NEXT and SKIP */
	    ctxt->nbChars ++;
	    ctxt->input->cur++;
	}
    } else {
	ctxt->errNo = XML_ERR_INVALID_CHARREF;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData,
	       "xmlParseCharRef: invalid value\n");
	ctxt->wellFormed = 0;
	ctxt->disableSAX = 1;
    }

    /*
     * [ WFC: Legal Character ]
     * Characters referred to using character references must match the
     * production for Char. 
     */
    if (IS_CHAR(val)) {
        return(val);
    } else {
	ctxt->errNo = XML_ERR_INVALID_CHAR;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData, "CharRef: invalid xmlChar value %d\n",
	                     val);
	ctxt->wellFormed = 0;
	ctxt->disableSAX = 1;
    }
    return(0);
}

/**
 * xmlParseStringCharRef:
 * @ctxt:  an XML parser context
 * @str:  a pointer to an index in the string
 *
 * parse Reference declarations, variant parsing from a string rather
 * than an an input flow.
 *
 * [66] CharRef ::= '&#' [0-9]+ ';' |
 *                  '&#x' [0-9a-fA-F]+ ';'
 *
 * [ WFC: Legal Character ]
 * Characters referred to using character references must match the
 * production for Char. 
 *
 * Returns the value parsed (as an int), 0 in case of error, str will be
 *         updated to the current value of the index
 */
int
xmlParseStringCharRef(xmlParserCtxtPtr ctxt, const xmlChar **str) {
    const xmlChar *ptr;
    xmlChar cur;
    int val = 0;

    if ((str == NULL) || (*str == NULL)) return(0);
    ptr = *str;
    cur = *ptr;
    if ((cur == '&') && (ptr[1] == '#') && (ptr[2] == 'x')) {
	ptr += 3;
	cur = *ptr;
	while (cur != ';') {
	    if ((cur >= '0') && (cur <= '9')) 
	        val = val * 16 + (cur - '0');
	    else if ((cur >= 'a') && (cur <= 'f'))
	        val = val * 16 + (cur - 'a') + 10;
	    else if ((cur >= 'A') && (cur <= 'F'))
	        val = val * 16 + (cur - 'A') + 10;
	    else {
		ctxt->errNo = XML_ERR_INVALID_HEX_CHARREF;
	        if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		    ctxt->sax->error(ctxt->userData, 
		         "xmlParseCharRef: invalid hexadecimal value\n");
		ctxt->wellFormed = 0;
		ctxt->disableSAX = 1;
		val = 0;
		break;
	    }
	    ptr++;
	    cur = *ptr;
	}
	if (cur == ';')
	    ptr++;
    } else if  ((cur == '&') && (ptr[1] == '#')){
	ptr += 2;
	cur = *ptr;
	while (cur != ';') {
	    if ((cur >= '0') && (cur <= '9')) 
	        val = val * 10 + (cur - '0');
	    else {
		ctxt->errNo = XML_ERR_INVALID_DEC_CHARREF;
	        if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		    ctxt->sax->error(ctxt->userData, 
		         "xmlParseCharRef: invalid decimal value\n");
		ctxt->wellFormed = 0;
		ctxt->disableSAX = 1;
		val = 0;
		break;
	    }
	    ptr++;
	    cur = *ptr;
	}
	if (cur == ';')
	    ptr++;
    } else {
	ctxt->errNo = XML_ERR_INVALID_CHARREF;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData,
	       "xmlParseCharRef: invalid value\n");
	ctxt->wellFormed = 0;
	ctxt->disableSAX = 1;
	return(0);
    }
    *str = ptr;

    /*
     * [ WFC: Legal Character ]
     * Characters referred to using character references must match the
     * production for Char. 
     */
    if (IS_CHAR(val)) {
        return(val);
    } else {
	ctxt->errNo = XML_ERR_INVALID_CHAR;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData,
		             "CharRef: invalid xmlChar value %d\n", val);
	ctxt->wellFormed = 0;
	ctxt->disableSAX = 1;
    }
    return(0);
}

/**
 * xmlParserHandleReference:
 * @ctxt:  the parser context
 * 
 * [67] Reference ::= EntityRef | CharRef
 *
 * [68] EntityRef ::= '&' Name ';'
 *
 * [ WFC: Entity Declared ]
 * the Name given in the entity reference must match that in an entity
 * declaration, except that well-formed documents need not declare any
 * of the following entities: amp, lt, gt, apos, quot. 
 *
 * [ WFC: Parsed Entity ]
 * An entity reference must not contain the name of an unparsed entity
 *
 * [66] CharRef ::= '&#' [0-9]+ ';' |
 *                  '&#x' [0-9a-fA-F]+ ';'
 *
 * A PEReference may have been detectect in the current input stream
 * the handling is done accordingly to 
 *      http://www.w3.org/TR/REC-xml#entproc
 */
void
xmlParserHandleReference(xmlParserCtxtPtr ctxt) {
    xmlParserInputPtr input;
    xmlChar *name;
    xmlEntityPtr ent = NULL;

    if (ctxt->token != 0) {
        return;
    }	
    if (RAW != '&') return;
    GROW;
    if ((RAW == '&') && (NXT(1) == '#')) {
	switch(ctxt->instate) {
	    case XML_PARSER_ENTITY_DECL:
	    case XML_PARSER_PI:
	    case XML_PARSER_CDATA_SECTION:
	    case XML_PARSER_COMMENT:
	    case XML_PARSER_SYSTEM_LITERAL:
		/* we just ignore it there */
		return;
	    case XML_PARSER_START_TAG:
		return;
	    case XML_PARSER_END_TAG:
		return;
	    case XML_PARSER_EOF:
		ctxt->errNo = XML_ERR_CHARREF_AT_EOF;
		if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		    ctxt->sax->error(ctxt->userData, "CharRef at EOF\n");
		ctxt->wellFormed = 0;
		ctxt->disableSAX = 1;
		return;
	    case XML_PARSER_PROLOG:
	    case XML_PARSER_START:
	    case XML_PARSER_MISC:
		ctxt->errNo = XML_ERR_CHARREF_IN_PROLOG;
		if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		    ctxt->sax->error(ctxt->userData, "CharRef in prolog!\n");
		ctxt->wellFormed = 0;
		ctxt->disableSAX = 1;
		return;
	    case XML_PARSER_EPILOG:
		ctxt->errNo = XML_ERR_CHARREF_IN_EPILOG;
		if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		    ctxt->sax->error(ctxt->userData, "CharRef in epilog!\n");
		ctxt->wellFormed = 0;
		ctxt->disableSAX = 1;
		return;
	    case XML_PARSER_DTD:
		ctxt->errNo = XML_ERR_CHARREF_IN_DTD;
		if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		    ctxt->sax->error(ctxt->userData, 
		           "CharRef are forbiden in DTDs!\n");
		ctxt->wellFormed = 0;
		ctxt->disableSAX = 1;
		return;
	    case XML_PARSER_ENTITY_VALUE:
	        /*
		 * NOTE: in the case of entity values, we don't do the
		 *       substitution here since we need the literal
		 *       entity value to be able to save the internal
		 *       subset of the document.
		 *       This will be handled by xmlDecodeEntities
		 */
		return;
	    case XML_PARSER_CONTENT:
	    case XML_PARSER_ATTRIBUTE_VALUE:
		ctxt->token = xmlParseCharRef(ctxt);
		return;
	}
	return;
    }

    switch(ctxt->instate) {
	case XML_PARSER_CDATA_SECTION:
	    return;
	case XML_PARSER_PI:
        case XML_PARSER_COMMENT:
	case XML_PARSER_SYSTEM_LITERAL:
        case XML_PARSER_CONTENT:
	    return;
	case XML_PARSER_START_TAG:
	    return;
	case XML_PARSER_END_TAG:
	    return;
        case XML_PARSER_EOF:
	    ctxt->errNo = XML_ERR_ENTITYREF_AT_EOF;
	    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	        ctxt->sax->error(ctxt->userData, "Reference at EOF\n");
	    ctxt->wellFormed = 0;
	    ctxt->disableSAX = 1;
	    return;
        case XML_PARSER_PROLOG:
	case XML_PARSER_START:
	case XML_PARSER_MISC:
	    ctxt->errNo = XML_ERR_ENTITYREF_IN_PROLOG;
	    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	        ctxt->sax->error(ctxt->userData, "Reference in prolog!\n");
	    ctxt->wellFormed = 0;
	    ctxt->disableSAX = 1;
	    return;
        case XML_PARSER_EPILOG:
	    ctxt->errNo = XML_ERR_ENTITYREF_IN_EPILOG;
	    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	        ctxt->sax->error(ctxt->userData, "Reference in epilog!\n");
	    ctxt->wellFormed = 0;
	    ctxt->disableSAX = 1;
	    return;
	case XML_PARSER_ENTITY_VALUE:
	    /*
	     * NOTE: in the case of entity values, we don't do the
	     *       substitution here since we need the literal
	     *       entity value to be able to save the internal
	     *       subset of the document.
	     *       This will be handled by xmlDecodeEntities
	     */
	    return;
        case XML_PARSER_ATTRIBUTE_VALUE:
	    /*
	     * NOTE: in the case of attributes values, we don't do the
	     *       substitution here unless we are in a mode where
	     *       the parser is explicitely asked to substitute
	     *       entities. The SAX callback is called with values
	     *       without entity substitution.
	     *       This will then be handled by xmlDecodeEntities
	     */
	    return;
	case XML_PARSER_ENTITY_DECL:
	    /*
	     * we just ignore it there
	     * the substitution will be done once the entity is referenced
	     */
	    return;
        case XML_PARSER_DTD:
	    ctxt->errNo = XML_ERR_ENTITYREF_IN_DTD;
	    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
		ctxt->sax->error(ctxt->userData, 
		       "Entity references are forbiden in DTDs!\n");
	    ctxt->wellFormed = 0;
	    ctxt->disableSAX = 1;
	    return;
    }

    NEXT;
    name = xmlScanName(ctxt);
    if (name == NULL) {
	ctxt->errNo = XML_ERR_ENTITYREF_NO_NAME;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData, "Entity reference: no name\n");
	ctxt->wellFormed = 0;
	ctxt->disableSAX = 1;
	ctxt->token = '&';
	return;
    }
    if (NXT(xmlStrlen(name)) != ';') {
	ctxt->errNo = XML_ERR_ENTITYREF_SEMICOL_MISSING;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData, 
	                     "Entity reference: ';' expected\n");
	ctxt->wellFormed = 0;
	ctxt->disableSAX = 1;
	ctxt->token = '&';
	xmlFree(name);
	return;
    }
    SKIP(xmlStrlen(name) + 1);
    if (ctxt->sax != NULL) {
	if (ctxt->sax->getEntity != NULL)
	    ent = ctxt->sax->getEntity(ctxt->userData, name);
    }

    /*
     * [ WFC: Entity Declared ]
     * the Name given in the entity reference must match that in an entity
     * declaration, except that well-formed documents need not declare any
     * of the following entities: amp, lt, gt, apos, quot. 
     */
    if (ent == NULL)
	ent = xmlGetPredefinedEntity(name);
    if (ent == NULL) {
        ctxt->errNo = XML_ERR_UNDECLARED_ENTITY;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData, 
			     "Entity reference: entity %s not declared\n",
			     name);
	ctxt->wellFormed = 0;
	ctxt->disableSAX = 1;
	xmlFree(name);
	return;
    }

    /*
     * [ WFC: Parsed Entity ]
     * An entity reference must not contain the name of an unparsed entity
     */
    if (ent->etype == XML_EXTERNAL_GENERAL_UNPARSED_ENTITY) {
        ctxt->errNo = XML_ERR_UNPARSED_ENTITY;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData, 
			 "Entity reference to unparsed entity %s\n", name);
	ctxt->wellFormed = 0;
	ctxt->disableSAX = 1;
    }

    if (ent->etype == XML_INTERNAL_PREDEFINED_ENTITY) {
        ctxt->token = ent->content[0];
	xmlFree(name);
	return;
    }
    input = xmlNewEntityInputStream(ctxt, ent);
    xmlPushInput(ctxt, input);
    xmlFree(name);
    return;
}

/**
 * xmlParserHandlePEReference:
 * @ctxt:  the parser context
 * 
 * [69] PEReference ::= '%' Name ';'
 *
 * [ WFC: No Recursion ]
 * TODO A parsed entity must not contain a recursive
 * reference to itself, either directly or indirectly. 
 *
 * [ WFC: Entity Declared ]
 * In a document without any DTD, a document with only an internal DTD
 * subset which contains no parameter entity references, or a document
 * with "standalone='yes'", ...  ... The declaration of a parameter
 * entity must precede any reference to it...
 *
 * [ VC: Entity Declared ]
 * In a document with an external subset or external parameter entities
 * with "standalone='no'", ...  ... The declaration of a parameter entity
 * must precede any reference to it...
 *
 * [ WFC: In DTD ]
 * Parameter-entity references may only appear in the DTD.
 * NOTE: misleading but this is handled.
 *
 * A PEReference may have been detected in the current input stream
 * the handling is done accordingly to 
 *      http://www.w3.org/TR/REC-xml#entproc
 * i.e. 
 *   - Included in literal in entity values
 *   - Included as Paraemeter Entity reference within DTDs
 */
void
xmlParserHandlePEReference(xmlParserCtxtPtr ctxt) {
    xmlChar *name;
    xmlEntityPtr entity = NULL;
    xmlParserInputPtr input;

    if (ctxt->token != 0) {
        return;
    }	
    if (RAW != '%') return;
    switch(ctxt->instate) {
	case XML_PARSER_CDATA_SECTION:
	    return;
        case XML_PARSER_COMMENT:
	    return;
	case XML_PARSER_START_TAG:
	    return;
	case XML_PARSER_END_TAG:
	    return;
        case XML_PARSER_EOF:
	    ctxt->errNo = XML_ERR_PEREF_AT_EOF;
	    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	        ctxt->sax->error(ctxt->userData, "PEReference at EOF\n");
	    ctxt->wellFormed = 0;
	    ctxt->disableSAX = 1;
	    return;
        case XML_PARSER_PROLOG:
	case XML_PARSER_START:
	case XML_PARSER_MISC:
	    ctxt->errNo = XML_ERR_PEREF_IN_PROLOG;
	    if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	        ctxt->sax->error(ctxt->userData, "PEReference in prolog!\n");
	    ctxt->wellFormed = 0;
	    ctxt->disableSAX = 1;
	    return;
	case XML_PARSER_ENTITY_DECL:
        case XML_PARSER_CONTENT: