more cleanup of the HTML parser to force it to not bypass SAX, Daniel.

Ready for 2.1.1 it seems

ChangeLog

@@ -1,3 +1,8 @@
+Fri Jun 30 20:29:08 MEST 2000
+
+	* HTMLparser.c HTMLtree.c SAX.c valid.c tree.h : more cleanup
+	  of the HTML parser to force it to not bypass SAX
+
 Fri Jun 30 11:19:59 CEST 2000 Daniel Veillard <Daniel.Veillard@w3.org>
 
 	* win32config.h.in: updated

HTMLparser.c

@@ -618,7 +618,7 @@ htmlAutoCloseOnClose(htmlParserCtxtPtr ctxt, const xmlChar *new) {
  */
 void
 htmlCheckImplied(htmlParserCtxtPtr ctxt, const xmlChar *new) {
-    if (!strcmp(new, "html"))
+    if (!xmlStrcmp(new, BAD_CAST"html"))
 	return;
     if (ctxt->nameNr <= 0) {
 #ifdef DEBUG
@@ -628,12 +628,15 @@ htmlCheckImplied(htmlParserCtxtPtr ctxt, const xmlChar *new) {
 	if ((ctxt->sax != NULL) && (ctxt->sax->startElement != NULL))
 	    ctxt->sax->startElement(ctxt->userData, BAD_CAST"html", NULL);
     }
-    if ((!strcmp(new, "body")) || (!strcmp(new, "head")))
+    if ((!xmlStrcmp(new, BAD_CAST"body")) || (!xmlStrcmp(new, BAD_CAST"head")))
         return;
     if (ctxt->nameNr <= 1) {
-	if ((!strcmp(new, "script")) || (!strcmp(new, "style")) ||
+	if ((!xmlStrcmp(new, BAD_CAST"script")) ||
-	    (!strcmp(new, "meta")) || (!strcmp(new, "link")) ||
+	    (!xmlStrcmp(new, BAD_CAST"style")) ||
-	    (!strcmp(new, "title")) || (!strcmp(new, "base"))) {
+	    (!xmlStrcmp(new, BAD_CAST"meta")) ||
+	    (!xmlStrcmp(new, BAD_CAST"link")) ||
+	    (!xmlStrcmp(new, BAD_CAST"title")) ||
+	    (!xmlStrcmp(new, BAD_CAST"base"))) {
 	    /* 
 	     * dropped OBJECT ... i you put it first BODY will be
 	     * assumed !
@@ -2152,17 +2155,15 @@ htmlParseDocTypeDecl(htmlParserCtxtPtr ctxt) {
 	    ctxt->sax->error(ctxt->userData, "DOCTYPE unproperly terminated\n");
 	ctxt->wellFormed = 0;
         /* We shouldn't try to resynchronize ... */
-    } else {
     }
     NEXT;
 
     /*
-     * Create the document accordingly to the DOCTYPE
+     * Create or update the document accordingly to the DOCTYPE
      */
-    if (ctxt->myDoc != NULL)
+    if ((ctxt->sax != NULL) && (ctxt->sax->internalSubset != NULL) &&
-        xmlFreeDoc(ctxt->myDoc);
+	(!ctxt->disableSAX))
-    
+	ctxt->sax->internalSubset(ctxt->userData, name, ExternalID, URI);
-    ctxt->myDoc = htmlNewDoc(URI, ExternalID);
 
     /*
      * Cleanup, since we don't use all those identifiers
@@ -2845,13 +2846,6 @@ htmlParseDocument(htmlParserCtxtPtr ctxt) {
     }
     SKIP_BLANKS;
 
-    /*
-     * Create the document if not done already.
-     */
-    if (ctxt->myDoc == NULL) {
-        ctxt->myDoc = htmlNewDoc(NULL, NULL);
-    }
-
     /*
      * Time to start parsing the tree itself
      */
@@ -3171,6 +3165,10 @@ htmlParseTryOrFinish(htmlParserCtxtPtr ctxt, int terminate) {
 		if ((ctxt->sax) && (ctxt->sax->setDocumentLocator))
 		    ctxt->sax->setDocumentLocator(ctxt->userData,
 						  &xmlDefaultSAXLocator);
+		if ((ctxt->sax) && (ctxt->sax->startDocument) &&
+	            (!ctxt->disableSAX))
+		    ctxt->sax->startDocument(ctxt->userData);
+
 		cur = in->cur[0];
 		next = in->cur[1];
 		if ((cur == '<') && (next == '!') &&
@@ -3190,7 +3188,6 @@ htmlParseTryOrFinish(htmlParserCtxtPtr ctxt, int terminate) {
 		    fprintf(stderr, "HPP: entering PROLOG\n");
 #endif
                 } else {
-		    ctxt->myDoc = htmlNewDoc(NULL, NULL);
 		    ctxt->instate = XML_PARSER_MISC;
 		}
 #ifdef DEBUG_PUSH

HTMLtree.c

@@ -158,6 +158,8 @@ htmlNodeDump(xmlBufferPtr buf, xmlDocPtr doc, xmlNodePtr cur) {
     /*
      * Special cases.
      */
+    if (cur->type == XML_DTD_NODE)
+	return;
     if (cur->type == XML_HTML_DOCUMENT_NODE) {
 	htmlDocContentDump(buf, (xmlDocPtr) cur);
 	return;

SAX.c

@@ -25,6 +25,7 @@
 #include <libxml/xmlIO.h>
 #include <libxml/SAX.h>
 #include <libxml/uri.h>
+#include <libxml/HTMLtree.h>
 
 /* #define DEBUG_SAX */
 /* #define DEBUG_SAX_TREE */
@@ -157,11 +158,22 @@ internalSubset(void *ctx, const xmlChar *name,
 	       const xmlChar *ExternalID, const xmlChar *SystemID)
 {
     xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) ctx;
+    xmlDtdPtr dtd;
 #ifdef DEBUG_SAX
     fprintf(stderr, "SAX.internalSubset(%s, %s, %s)\n",
             name, ExternalID, SystemID);
 #endif
-    xmlCreateIntSubset(ctxt->myDoc, name, ExternalID, SystemID);
+
+    if (ctxt->myDoc == NULL)
+	return;
+    dtd = xmlGetIntSubset(ctxt->myDoc);
+    if (dtd != NULL) {
+	xmlUnlinkNode((xmlNodePtr) dtd);
+	xmlFreeDtd(dtd);
+	ctxt->myDoc->intSubset = NULL;
+    }
+    ctxt->myDoc->intSubset = 
+	xmlCreateIntSubset(ctxt->myDoc, name, ExternalID, SystemID);
 }
 
 /**
@@ -1485,7 +1497,7 @@ xmlDefaultSAXHandlerInit(void)
  * Default handler for HTML, builds the DOM tree
  */
 xmlSAXHandler htmlDefaultSAXHandler = {
-    NULL,
+    internalSubset,
     NULL,
     NULL,
     NULL,
@@ -1522,7 +1534,7 @@ xmlSAXHandler htmlDefaultSAXHandler = {
 void
 htmlDefaultSAXHandlerInit(void)
 {
-    htmlDefaultSAXHandler.internalSubset = NULL;
+    htmlDefaultSAXHandler.internalSubset = internalSubset;
     htmlDefaultSAXHandler.externalSubset = NULL;
     htmlDefaultSAXHandler.isStandalone = NULL;
     htmlDefaultSAXHandler.hasInternalSubset = NULL;

doc/upgrade.html

@@ -48,7 +48,7 @@ mail</a>:</p>
     Use <strong>xmlDocGetRootElement(doc)</strong> to get the root element of
     a document. Alternatively if you are sure to not reference Dtds nor have
     PIs or comments before or after the root element s/->root/->children/g
-    will probably do it.    </li>
+    will probably do it.</li>
   <li>The white space issue, this one is more complex, unless special case of
     validating parsing, the line breaks and spaces usually used for indenting
     and formatting the document content becomes significant. So they are
@@ -90,7 +90,7 @@ They offers the following:</p>
     <strong>#include&lt;libxml/...></strong> in both cases.</li>
   <li>similar identifiers defined via macros for the child and root fields:
     respectively <strong>xmlChildrenNode</strong> and
-    <strong>xmlRootNode</strong> </li>
+    <strong>xmlRootNode</strong></li>
   <li>a new macro <strong>LIBXML_TEST_VERSION</strong> which should be
     inserted once in the client code</li>
 </ol>
@@ -118,7 +118,7 @@ following:</p>
     <strong>LIBXML_TEST_VERSION</strong> is a fine place).</li>
 </ol>
 
-<p>Following those 3 steps should work. It worked for some of my own code.</p>
+<p>Following those steps should work. It worked for some of my own code.</p>
 
 <p>Let me put some emphasis on the fact that there is far more changes from
 libxml 1.x to 2.x than the ones you may have to patch for. The overall code
@@ -128,6 +128,6 @@ upgrade, it may cost a lot on the long term ...</p>
 
 <p><a href="mailto:Daniel.Veillard@w3.org">Daniel Veillard</a></p>
 
-<p>$Id: upgrade.html,v 1.5 2000/05/06 08:11:18 veillard Exp $</p>
+<p>$Id: upgrade.html,v 1.6 2000/06/29 00:43:26 veillard Exp $</p>
 </body>
 </html>

include/libxml/tree.h

@@ -414,6 +414,7 @@ xmlDtdPtr	xmlNewDtd		(xmlDocPtr doc,
 					 const xmlChar *name,
 					 const xmlChar *ExternalID,
 					 const xmlChar *SystemID);
+xmlDtdPtr	xmlGetIntSubset		(xmlDocPtr doc);
 void		xmlFreeDtd		(xmlDtdPtr cur);
 xmlNsPtr	xmlNewGlobalNs		(xmlDocPtr doc,
 					 const xmlChar *href,

tree.h

@@ -414,6 +414,7 @@ xmlDtdPtr	xmlNewDtd		(xmlDocPtr doc,
 					 const xmlChar *name,
 					 const xmlChar *ExternalID,
 					 const xmlChar *SystemID);
+xmlDtdPtr	xmlGetIntSubset		(xmlDocPtr doc);
 void		xmlFreeDtd		(xmlDtdPtr cur);
 xmlNsPtr	xmlNewGlobalNs		(xmlDocPtr doc,
 					 const xmlChar *href,

valid.c

@@ -2031,6 +2031,9 @@ xmlIsRef(xmlDocPtr doc, xmlNodePtr elem, xmlAttrPtr attr) {
             ((attr->name[1] == 'D') || (attr->name[1] == 'd')) &&
 	    (attr->name[2] == 0)) return(1);
 	 *******************/
+    } else if (doc->type == XML_HTML_DOCUMENT_NODE) {
+	/* TODO @@@ */
+	return(0);    
     } else {
 	xmlAttributePtr attrDecl;