htmlparser.c

SIP 1.5.0源代码

 * Returns the new input stream or NULL
 */
htmlParserInputPtr
htmlNewInputStream(htmlParserCtxtPtr ctxt) {
    htmlParserInputPtr input;

    input = (xmlParserInputPtr) xmlMalloc(sizeof(htmlParserInput));
    if (input == NULL) {
        ctxt->errNo = XML_ERR_NO_MEMORY;
	if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData, 
	                     "malloc: couldn't allocate a new input stream\n");
	ctxt->errNo = XML_ERR_NO_MEMORY;
	return(NULL);
    }
    input->filename = NULL;
    input->directory = NULL;
    input->base = NULL;
    input->cur = NULL;
    input->buf = NULL;
    input->line = 1;
    input->col = 1;
    input->buf = NULL;
    input->free = NULL;
    input->consumed = 0;
    input->length = 0;
    return(input);
}


/************************************************************************
 *									*
 *		Commodity functions, cleanup needed ?			*
 *									*
 ************************************************************************/

/**
 * areBlanks:
 * @ctxt:  an HTML parser context
 * @str:  a xmlChar *
 * @len:  the size of @str
 *
 * Is this a sequence of blank chars that one can ignore ?
 *
 * Returns 1 if ignorable 0 otherwise.
 */

static int areBlanks(htmlParserCtxtPtr ctxt, const xmlChar *str, int len) {
    int i;
    xmlNodePtr lastChild;

    for (i = 0;i < len;i++)
        if (!(IS_BLANK(str[i]))) return(0);

    if (CUR != '<') return(0);
    if (ctxt->node == NULL) return(0);
    lastChild = xmlGetLastChild(ctxt->node);
    if (lastChild == NULL) {
        if (ctxt->node->content != NULL) return(0);
    } else if (xmlNodeIsText(lastChild))
        return(0);
    return(1);
}

/**
 * htmlHandleEntity:
 * @ctxt:  an HTML parser context
 * @entity:  an XML entity pointer.
 *
 * Default handling of an HTML entity, call the parser with the
 * substitution string
 */

void
htmlHandleEntity(htmlParserCtxtPtr ctxt, xmlEntityPtr entity) {
    int len;

    if (entity->content == NULL) {
        if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL))
	    ctxt->sax->error(ctxt->userData, "htmlHandleEntity %s: content == NULL\n",
	               entity->name);
	ctxt->wellFormed = 0;
        return;
    }
    len = xmlStrlen(entity->content);

    /*
     * Just handle the content as a set of chars.
     */
    if ((ctxt->sax != NULL) && (ctxt->sax->characters != NULL))
	ctxt->sax->characters(ctxt->userData, entity->content, len);

}

/**
 * htmlNewDoc:
 * @URI:  URI for the dtd, or NULL
 * @ExternalID:  the external ID of the DTD, or NULL
 *
 * Returns a new document
 */
htmlDocPtr
htmlNewDoc(const xmlChar *URI, const xmlChar *ExternalID) {
    xmlDocPtr cur;

    /*
     * Allocate a new document and fill the fields.
     */
    cur = (xmlDocPtr) xmlMalloc(sizeof(xmlDoc));
    if (cur == NULL) {
        fprintf(stderr, "xmlNewDoc : malloc failed\n");
	return(NULL);
    }
    memset(cur, 0, sizeof(xmlDoc));

    cur->type = XML_HTML_DOCUMENT_NODE;
    cur->version = NULL;
    cur->intSubset = NULL;
    if ((ExternalID == NULL) &&
	(URI == NULL))
	xmlCreateIntSubset(cur, BAD_CAST "HTML",
		    BAD_CAST "-//W3C//DTD HTML 4.0 Transitional//EN",
		    BAD_CAST "http://www.w3.org/TR/REC-html40/loose.dtd");
    else
	xmlCreateIntSubset(cur, BAD_CAST "HTML", ExternalID, URI);
    cur->name = NULL;
    cur->children = NULL; 
    cur->extSubset = NULL;
    cur->oldNs = NULL;
    cur->encoding = NULL;
    cur->standalone = 1;
    cur->compression = 0;
    cur->ids = NULL;
    cur->refs = NULL;
#ifndef XML_WITHOUT_CORBA
    cur->_private = NULL;
#endif
    return(cur);
}


/************************************************************************
 *									*
 *			The parser itself				*
 *	Relates to http://www.w3.org/TR/html40				*
 *									*
 ************************************************************************/

/************************************************************************
 *									*
 *			The parser itself				*
 *									*
 ************************************************************************/

/**
 * htmlParseHTMLName:
 * @ctxt:  an HTML parser context
 *
 * parse an HTML tag or attribute name, note that we convert it to lowercase
 * since HTML names are not case-sensitive.
 *
 * Returns the Tag Name parsed or NULL
 */

xmlChar *
htmlParseHTMLName(htmlParserCtxtPtr ctxt) {
    xmlChar *ret = NULL;
    int i = 0;
    xmlChar loc[HTML_PARSER_BUFFER_SIZE];

    if (!IS_LETTER(CUR) && (CUR != '_') &&
        (CUR != ':')) return(NULL);

    while ((i < HTML_PARSER_BUFFER_SIZE) &&
           ((IS_LETTER(CUR)) || (IS_DIGIT(CUR)))) {
	if ((CUR >= 'A') && (CUR <= 'Z')) loc[i] = CUR + 0x20;
        else loc[i] = CUR;
	i++;
	
	NEXT;
    }
    
    ret = xmlStrndup(loc, i);

    return(ret);
}

/**
 * htmlParseName:
 * @ctxt:  an HTML parser context
 *
 * parse an HTML name, this routine is case sensistive.
 *
 * Returns the Name parsed or NULL
 */

xmlChar *
htmlParseName(htmlParserCtxtPtr ctxt) {
    xmlChar buf[HTML_MAX_NAMELEN];
    int len = 0;

    GROW;
    if (!IS_LETTER(CUR) && (CUR != '_')) {
	return(NULL);
    }

    while ((IS_LETTER(CUR)) || (IS_DIGIT(CUR)) ||
           (CUR == '.') || (CUR == '-') ||
	   (CUR == '_') || (CUR == ':') || 
	   (IS_COMBINING(CUR)) ||
	   (IS_EXTENDER(CUR))) {
	buf[len++] = CUR;
	NEXT;
	if (len >= HTML_MAX_NAMELEN) {
	    fprintf(stderr, 
	       "htmlParseName: reached HTML_MAX_NAMELEN limit\n");
	    while ((IS_LETTER(CUR)) || (IS_DIGIT(CUR)) ||
		   (CUR == '.') || (CUR == '-') ||
		   (CUR == '_') || (CUR == ':') || 
		   (IS_COMBINING(CUR)) ||
		   (IS_EXTENDER(CUR)))
		 NEXT;
	    break;
	}
    }
    return(xmlStrndup(buf, len));
}

/**
 * htmlParseHTMLAttribute:
 * @ctxt:  an HTML parser context
 * @stop:  a char stop value
 * 
 * parse an HTML attribute value till the stop (quote), if
 * stop is 0 then it stops at the first space
 *
 * Returns the attribute parsed or NULL
 */

xmlChar *
htmlParseHTMLAttribute(htmlParserCtxtPtr ctxt, const xmlChar stop) {
#if 0
    xmlChar buf[HTML_MAX_NAMELEN];
    int len = 0;

    GROW;
    while ((CUR != 0) && (CUR != stop) && (CUR != '>')) {
	if ((stop == 0) && (IS_BLANK(CUR))) break;
	buf[len++] = CUR;
	NEXT;
	if (len >= HTML_MAX_NAMELEN) {
	    fprintf(stderr, 
	       "htmlParseHTMLAttribute: reached HTML_MAX_NAMELEN limit\n");
	    while ((!IS_BLANK(CUR)) && (CUR != '<') &&
		   (CUR != '>') &&
		   (CUR != '\'') && (CUR != '"'))
		 NEXT;
	    break;
	}
    }
    return(xmlStrndup(buf, len));
#else    
    xmlChar *buffer = NULL;
    int buffer_size = 0;
    xmlChar *out = NULL;
    xmlChar *name = NULL;

    xmlChar *cur = NULL;
    htmlEntityDescPtr ent;

    /*
     * allocate a translation buffer.
     */
    buffer_size = HTML_PARSER_BIG_BUFFER_SIZE;
    buffer = (xmlChar *) xmlMalloc(buffer_size * sizeof(xmlChar));
    if (buffer == NULL) {
	perror("htmlParseHTMLAttribute: malloc failed");
	return(NULL);
    }
    out = buffer;

    /*
     * Ok loop until we reach one of the ending chars
     */
    while ((CUR != 0) && (CUR != stop) && (CUR != '>')) {
	if ((stop == 0) && (IS_BLANK(CUR))) break;
        if (CUR == '&') {
	    if (NXT(1) == '#') {
		int val = htmlParseCharRef(ctxt);
		*out++ = val;
	    } else {
		ent = htmlParseEntityRef(ctxt, &name);
		if (name == NULL) {
		    *out++ = '&';
		    if (out - buffer > buffer_size - 100) {
			int index = out - buffer;

			growBuffer(buffer);
			out = &buffer[index];
		    }
		} else if ((ent == NULL) || (ent->value <= 0) ||
		           (ent->value >= 255)) {
		    *out++ = '&';
		    cur = name;
		    while (*cur != 0) {
			if (out - buffer > buffer_size - 100) {
			    int index = out - buffer;

			    growBuffer(buffer);
			    out = &buffer[index];
			}
			*out++ = *cur++;
		    }
		    xmlFree(name);
		} else {
		    *out++ = ent->value;
		    if (out - buffer > buffer_size - 100) {
			int index = out - buffer;

			growBuffer(buffer);
			out = &buffer[index];
		    }
		    xmlFree(name);
		}
	    }
	} else {
	    *out++ = CUR;
	    if (out - buffer > buffer_size - 100) {
	      int index = out - buffer;
	      
	      growBuffer(buffer);
	      out = &buffer[index];
	    }
	    NEXT;
	}
    }
    *out++ = 0;
    return(buffer);
#endif
}

/**
 * htmlParseNmtoken:
 * @ctxt:  an HTML parser context
 * 
 * parse an HTML Nmtoken.
 *
 * Returns the Nmtoken parsed or NULL
 */

xmlChar *
htmlParseNmtoken(htmlParserCtxtPtr ctxt) {
    xmlChar buf[HTML_MAX_NAMELEN];
    int len = 0;

    GROW;
    while ((IS_LETTER(CUR)) || (IS_DIGIT(CUR)) ||
           (CUR == '.') || (CUR == '-') ||
	   (CUR == '_') || (CUR == ':') || 
	   (IS_COMBINING(CUR)) ||
	   (IS_EXTENDER(CUR))) {
	buf[len++] = CUR;
	NEXT;
	if (len >= HTML_MAX_NAMELEN) {
	    fprintf(stderr, 
	       "htmlParseNmtoken: reached HTML_MAX_NAMELEN limit\n");
	    while ((IS_LETTER(CUR)) || (IS_DIGIT(CUR)) ||
		   (CUR == '.') || (CUR == '-') ||
		   (CUR == '_') || (CUR == ':') || 
		   (IS_COMBINING(CUR)) ||
		   (IS_EXTENDER(CUR)))
		 NEXT;
	    break;
	}
    }
    return(xmlStrndup(buf, len));
}

/**
 * htmlParseEntityRef:
 * @ctxt:  an HTML parser context
 * @str:  location to store the entity name
 *
 * parse an HTML ENTITY references
 *
 * [68] EntityRef ::= '&' Name ';'
 *
 * Returns the associated htmlEntityDescPtr if found, or NULL otherwise,
 *         if non-NULL *str will have to be freed by the caller.
 */
htmlEntityDescPtr
htmlParseEntityRef(htmlParserCtxtPtr ctxt, xmlChar **str) {
    xmlChar *name;
    htmlEntityDescPtr ent = NULL;
    *str = NULL;

    if (CUR == '&') {
        NEXT;
        name = htmlParseName(ctxt);
	if (name == NULL) {