fix various attribute normalisation problems reported by Ashwin this

* parser.c: fix various attribute normalisation problems reported
  by Ashwin
* result/c14n/without-comments/example-4
  result/c14n/with-comments/example-4: this impacted the result of
  two c14n tests :-\
* test/att9 test/att10 test/att11 result//att9* result//att10*
  result//att11*: added 3 specific regression tests coming from the
  XML spec revision and from Ashwin
Daniel

svn path=/trunk/; revision=3715

ChangeLog

@@ -1,3 +1,14 @@
+Tue Mar 25 17:48:02 CET 2008 Daniel Veillard <daniel@veillard.com>
+
+	* parser.c: fix various attribute normalisation problems reported
+	  by Ashwin
+	* result/c14n/without-comments/example-4
+	  result/c14n/with-comments/example-4: this impacted the result of
+	  two c14n tests :-\
+	* test/att9 test/att10 test/att11 result//att9* result//att10*
+	  result//att11*: added 3 specific regression tests coming from the
+	  XML spec revision and from Ashwin
+
 Tue Mar 25 14:20:49 CET 2008 Daniel Veillard <daniel@veillard.com>
 
 	* uri.c: fix saving for file:///X:/ URI embedding Windows file paths

debugXML.c

@@ -320,7 +320,8 @@ xmlCtxtGenericNodeCheck(xmlDebugCtxtPtr ctxt, xmlNodePtr node) {
     }
     if (node->next == NULL) {
 	if ((node->parent != NULL) && (node->type != XML_ATTRIBUTE_NODE) &&
-	    (node->parent->last != node))
+	    (node->parent->last != node) &&
+	    (node->parent->type == XML_ELEMENT_NODE))
 	    xmlDebugErr(ctxt, XML_CHECK_NO_NEXT,
                     "Node has no next and not last of parent list\n");
     } else {

parser.c

@@ -862,6 +862,103 @@ struct _xmlDefAttrs {
     const xmlChar *values[4]; /* array of localname/prefix/values */
 };
 
+/**
+ * xmlAttrNormalizeSpace:
+ * @src: the source string
+ * @dst: the target string
+ *
+ * Normalize the space in non CDATA attribute values:
+ * If the attribute type is not CDATA, then the XML processor MUST further
+ * process the normalized attribute value by discarding any leading and
+ * trailing space (#x20) characters, and by replacing sequences of space
+ * (#x20) characters by a single space (#x20) character.
+ * Note that the size of dst need to be at least src, and if one doesn't need
+ * to preserve dst (and it doesn't come from a dictionary or read-only) then
+ * passing src as dst is just fine.
+ *
+ * Returns a pointer to the normalized value (dst) or NULL if no conversion
+ *         is needed.
+ */
+static xmlChar *
+xmlAttrNormalizeSpace(const xmlChar *src, xmlChar *dst)
+{
+    if ((src == NULL) || (dst == NULL))
+        return(NULL);
+
+    while (*src == 0x20) src++;
+    while (*src != 0) {
+	if (*src == 0x20) {
+	    while (*src == 0x20) src++;
+	    if (*src != 0)
+		*dst++ = 0x20;
+	} else {
+	    *dst++ = *src++;
+	}
+    }
+    *dst = 0;
+    if (dst == src)
+       return(NULL);
+    return(dst);
+}
+
+/**
+ * xmlAttrNormalizeSpace2:
+ * @src: the source string
+ *
+ * Normalize the space in non CDATA attribute values, a slightly more complex
+ * front end to avoid allocation problems when running on attribute values
+ * coming from the input.
+ *
+ * Returns a pointer to the normalized value (dst) or NULL if no conversion
+ *         is needed.
+ */
+static const xmlChar *
+xmlAttrNormalizeSpace2(xmlParserCtxtPtr ctxt, const xmlChar *src, int *len)
+{
+    int i;
+    int remove_head = 0;
+    int need_realloc = 0;
+    const xmlChar *cur;
+
+    if ((ctxt == NULL) || (src == NULL) || (len == NULL))
+        return(NULL);
+    i = *len;
+    if (i <= 0)
+        return(NULL);
+
+    cur = src;
+    while (*cur == 0x20) {
+        cur++;
+	remove_head++;
+    }
+    while (*cur != 0) {
+	if (*cur == 0x20) {
+	    cur++;
+	    if ((*cur == 0x20) || (*cur == 0)) {
+	        need_realloc = 1;
+		break;
+	    }
+	} else
+	    cur++;
+    }
+    if (need_realloc) {
+        xmlChar *ret;
+
+	ret = xmlStrndup(src + remove_head, i - remove_head + 1);
+	if (ret == NULL) {
+	    xmlErrMemory(ctxt, NULL);
+	    return(NULL);
+	}
+	xmlAttrNormalizeSpace(ret, ret);
+	*len = (int) strlen((const char *)ret);
+        return(ret);
+    } else if (remove_head) {
+        *len -= remove_head;
+	return(src + remove_head);
+    }
+    return(NULL);
+}
+
 /**
  * xmlAddDefAttrs:
  * @ctxt:  an XML parser context
@@ -5020,6 +5117,8 @@ xmlParseAttributeListDecl(xmlParserCtxtPtr ctxt) {
 		    xmlFreeEnumeration(tree);
 	        break;
 	    }
+	    if ((type != XML_ATTRIBUTE_CDATA) && (defaultValue != NULL))
+	        xmlAttrNormalizeSpace(defaultValue, defaultValue);
 
 	    GROW;
             if (RAW != '>') {
@@ -7900,9 +7999,10 @@ need_complex:
 
 static const xmlChar *
 xmlParseAttribute2(xmlParserCtxtPtr ctxt,
-                   const xmlChar *pref, const xmlChar *elem,
+                   const xmlChar * pref, const xmlChar * elem,
-                   const xmlChar **prefix, xmlChar **value,
+                   const xmlChar ** prefix, xmlChar ** value,
-		   int *len, int *alloc) {
+                   int *len, int *alloc)
+{
     const xmlChar *name;
     xmlChar *val, *internal_val = NULL;
     int normalize = 0;
@@ -7913,7 +8013,7 @@ xmlParseAttribute2(xmlParserCtxtPtr ctxt,
     if (name == NULL) {
         xmlFatalErrMsg(ctxt, XML_ERR_NAME_REQUIRED,
                        "error parsing attribute name\n");
-        return(NULL);
+        return (NULL);
     }
 
     /*
@@ -7924,7 +8024,8 @@ xmlParseAttribute2(xmlParserCtxtPtr ctxt,
 
         type = (int) (long) xmlHashQLookup2(ctxt->attsSpecial,
                                             pref, elem, *prefix, name);
-	if (type != 0) normalize = 1;
+        if (type != 0)
+            normalize = 1;
     }
 
     /*
@@ -7935,11 +8036,29 @@ xmlParseAttribute2(xmlParserCtxtPtr ctxt,
         NEXT;
         SKIP_BLANKS;
         val = xmlParseAttValueInternal(ctxt, len, alloc, normalize);
+	if (normalize) {
+	    /*
+	     * Sometimes a second normalisation pass for spaces is needed
+	     * but that only happens if charrefs or entities refernces
+	     * have been used in the attribute value, i.e. the attribute
+	     * value have been extracted in an allocated string already.
+	     */
+	    if (*alloc) {
+	        const xmlChar *val2;
+
+	        val2 = xmlAttrNormalizeSpace2(ctxt, val, len);
+		if (val2 != NULL) {
+		    xmlFree(val);
+		    val = val2;
+		}
+	    }
+	}
         ctxt->instate = XML_PARSER_CONTENT;
     } else {
         xmlFatalErrMsgStr(ctxt, XML_ERR_ATTRIBUTE_WITHOUT_VALUE,
-	       "Specification mandate value for attribute %s\n", name);
+                          "Specification mandate value for attribute %s\n",
-	return(NULL);
+                          name);
+        return (NULL);
     }
 
     if (*prefix == ctxt->str_xml) {
@@ -7968,7 +8087,7 @@ xmlParseAttribute2(xmlParserCtxtPtr ctxt,
                 *(ctxt->space) = 1;
             else {
                 xmlWarningMsg(ctxt, XML_WAR_SPACE_VALUE,
-"Invalid value \"%s\" for xml:space : \"default\" or \"preserve\" expected\n",
+                              "Invalid value \"%s\" for xml:space : \"default\" or \"preserve\" expected\n",
                               internal_val, NULL);
             }
         }
@@ -7978,9 +8097,8 @@ xmlParseAttribute2(xmlParserCtxtPtr ctxt,
     }
 
     *value = val;
-    return(name);
+    return (name);
 }
-
 /**
  * xmlParseStartTag2:
  * @ctxt:  an XML parser context

result/att10

@@ -0,0 +1,18 @@
+<?xml version="1.0"?>
+<!DOCTYPE doc [
+<!ELEMENT doc (tst)*>
+<!ELEMENT tst (#PCDATA)>
+<!ATTLIST tst a NMTOKENS #IMPLIED>
+<!ATTLIST tst b CDATA #IMPLIED>
+<!ENTITY d "&#xD;">
+<!ENTITY a "&#xA;">
+<!ENTITY da "&#xD;&#xA;">
+]>
+<doc>
+<tst a="xyz" b="  xyz"/>
+<tst a="&d;&d;A&a; &a;B&da;" b="&d;&d;A&a; &a;B&da;"/>
+<tst a="&#13;&#13;A&#10;&#10;B&#13;&#10;" b="&#13;&#13;A&#10;&#10;B&#13;&#10;"/>
+<tst a="x y" b=" x  y "/>
+<tst a="a b" b=" a b "/>
+<tst a="a b" b="  a  b  "/>
+</doc>

result/att10.rde

@@ -0,0 +1,23 @@
+0 10 doc 0 0
+0 1 doc 0 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+0 15 doc 0 0

result/att10.rdr

@@ -0,0 +1,23 @@
+0 10 doc 0 0
+0 1 doc 0 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+1 1 tst 1 0
+1 14 #text 0 1 
+
+0 15 doc 0 0

result/att10.sax

@@ -0,0 +1,61 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.internalSubset(doc, , )
+SAX.elementDecl(doc, 4, ...)
+SAX.elementDecl(tst, 3, ...)
+SAX.attributeDecl(tst, a, 8, 3, NULL, ...)
+SAX.attributeDecl(tst, b, 1, 3, NULL, ...)
+SAX.entityDecl(d, 1, (null), (null), 
+)
+SAX.getEntity(d)
+SAX.entityDecl(a, 1, (null), (null), 
+)
+SAX.getEntity(a)
+SAX.entityDecl(da, 1, (null), (null), 
+)
+SAX.getEntity(da)
+SAX.externalSubset(doc, , )
+SAX.startElement(doc)
+SAX.characters(
+, 1)
+SAX.startElement(tst, a='  xyz', b='  xyz')
+SAX.endElement(tst)
+SAX.characters(
+, 1)
+SAX.getEntity(d)
+SAX.getEntity(d)
+SAX.getEntity(a)
+SAX.getEntity(a)
+SAX.getEntity(da)
+SAX.getEntity(d)
+SAX.getEntity(d)
+SAX.getEntity(a)
+SAX.getEntity(a)
+SAX.getEntity(da)
+SAX.startElement(tst, a='&d;&d;A&a; &a;B&da;', b='&d;&d;A&a; &a;B&da;')
+SAX.endElement(tst)
+SAX.characters(
+, 1)
+SAX.startElement(tst, a='
+
+A
+
+B
+', b='
+
+A
+
+B
+')
+SAX.endElement(tst)
+SAX.characters(
+, 1)
+SAX.startElement(tst, a=' x  y ', b=' x  y ')
+SAX.endElement(tst)
+SAX.characters(
+, 1)
+SAX.startElement(tst, a=' a b ', b=' a b ')
+SAX.endElement(tst)
+SAX.characters(
+, 1)
+SAX.startElement(tst, a='  a  b  ', b='  a  b  ')

result/att10.sax2

@@ -0,0 +1,57 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.internalSubset(doc, , )
+SAX.elementDecl(doc, 4, ...)
+SAX.elementDecl(tst, 3, ...)
+SAX.attributeDecl(tst, a, 8, 3, NULL, ...)
+SAX.attributeDecl(tst, b, 1, 3, NULL, ...)
+SAX.entityDecl(d, 1, (null), (null), 
+)
+SAX.getEntity(d)
+SAX.entityDecl(a, 1, (null), (null), 
+)
+SAX.getEntity(a)
+SAX.entityDecl(da, 1, (null), (null), 
+)
+SAX.getEntity(da)
+SAX.externalSubset(doc, , )
+SAX.startElementNs(doc, NULL, NULL, 0, 0, 0)
+SAX.characters(
+, 1)
+SAX.startElementNs(tst, NULL, NULL, 0, 2, 0, a='xyz"...', 3, b='  xy...', 5)
+SAX.endElementNs(tst, NULL, NULL)
+SAX.characters(
+, 1)
+SAX.getEntity(d)
+SAX.getEntity(d)
+SAX.getEntity(a)
+SAX.getEntity(a)
+SAX.getEntity(da)
+SAX.getEntity(d)
+SAX.getEntity(d)
+SAX.getEntity(a)
+SAX.getEntity(a)
+SAX.getEntity(da)
+SAX.startElementNs(tst, NULL, NULL, 0, 2, 0, a='&d;&...', 19, b='&d;&...', 19)
+SAX.endElementNs(tst, NULL, NULL)
+SAX.characters(
+, 1)
+SAX.startElementNs(tst, NULL, NULL, 0, 2, 0, a='
+
+A
+...', 8, b='
+
+A
+...', 8)
+SAX.endElementNs(tst, NULL, NULL)
+SAX.characters(
+, 1)
+SAX.startElementNs(tst, NULL, NULL, 0, 2, 0, a='x y...', 3, b=' x  ...', 6)
+SAX.endElementNs(tst, NULL, NULL)
+SAX.characters(
+, 1)
+SAX.startElementNs(tst, NULL, NULL, 0, 2, 0, a='a b ...', 3, b=' a b...', 5)
+SAX.endElementNs(tst, NULL, NULL)
+SAX.characters(
+, 1)
+SAX.startElementNs(tst, NULL, NULL, 0, 2, 0, a='a b...', 3, b='  a ...', 8)

result/att11

@@ -0,0 +1,9 @@
+<?xml version="1.0" standalone="yes"?>
+<!DOCTYPE attributes [
+<!ELEMENT attributes EMPTY>
+<!ATTLIST attributes nmtoken NMTOKEN #IMPLIED>
+<!ATTLIST attributes nmtokens NMTOKENS #IMPLIED>
+<!ENTITY ent " entity&recursive; ">
+<!ENTITY recursive "reference">
+]>
+<attributes nmtoken="&ent; &ent; &ent;" nmtokens="Test&#13;&#10; this normalization"/>

result/att11.rde

@@ -0,0 +1,2 @@
+0 10 attributes 0 0
+0 1 attributes 1 0

result/att11.rdr

@@ -0,0 +1,2 @@
+0 10 attributes 0 0
+0 1 attributes 1 0

result/att11.sax

@@ -0,0 +1,21 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.internalSubset(attributes, , )
+SAX.elementDecl(attributes, 1, ...)
+SAX.attributeDecl(attributes, nmtoken, 7, 3, NULL, ...)
+SAX.attributeDecl(attributes, nmtokens, 8, 3, NULL, ...)
+SAX.entityDecl(ent, 1, (null), (null),  entity&recursive; )
+SAX.getEntity(ent)
+SAX.entityDecl(recursive, 1, (null), (null), reference)
+SAX.getEntity(recursive)
+SAX.externalSubset(attributes, , )
+SAX.getEntity(ent)
+SAX.getEntity(recursive)
+SAX.getEntity(ent)
+SAX.getEntity(recursive)
+SAX.getEntity(ent)
+SAX.getEntity(recursive)
+SAX.startElement(attributes, nmtoken=' &ent;   &ent; &ent; ', nmtokens=' Test
+     this  normalization ')
+SAX.endElement(attributes)
+SAX.endDocument()

result/att11.sax2

@@ -0,0 +1,20 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.internalSubset(attributes, , )
+SAX.elementDecl(attributes, 1, ...)
+SAX.attributeDecl(attributes, nmtoken, 7, 3, NULL, ...)
+SAX.attributeDecl(attributes, nmtokens, 8, 3, NULL, ...)
+SAX.entityDecl(ent, 1, (null), (null),  entity&recursive; )
+SAX.getEntity(ent)
+SAX.entityDecl(recursive, 1, (null), (null), reference)
+SAX.getEntity(recursive)
+SAX.externalSubset(attributes, , )
+SAX.getEntity(ent)
+SAX.getEntity(recursive)
+SAX.getEntity(ent)
+SAX.getEntity(recursive)
+SAX.getEntity(ent)
+SAX.getEntity(recursive)
+SAX.startElementNs(attributes, NULL, NULL, 0, 2, 0, nmtoken='&ent...', 17, nmtokens='Test...', 25)
+SAX.endElementNs(attributes, NULL, NULL)
+SAX.endDocument()

result/att9

@@ -0,0 +1,6 @@
+<?xml version="1.0"?>
+<!DOCTYPE doc [
+<!ATTLIST doc a1 NMTOKENS "1 2">
+<!ELEMENT doc (#PCDATA)>
+]>
+<doc/>

result/att9.rde

@@ -0,0 +1,2 @@
+0 10 doc 0 0
+0 1 doc 1 0

result/att9.rdr

@@ -0,0 +1,2 @@
+0 10 doc 0 0
+0 1 doc 1 0

result/att9.sax

@@ -0,0 +1,9 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.internalSubset(doc, , )
+SAX.attributeDecl(doc, a1, 8, 1, 1 2, ...)
+SAX.elementDecl(doc, 3, ...)
+SAX.externalSubset(doc, , )
+SAX.startElement(doc)
+SAX.endElement(doc)
+SAX.endDocument()

result/att9.sax2

@@ -0,0 +1,9 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.internalSubset(doc, , )
+SAX.attributeDecl(doc, a1, 8, 1, 1 2, ...)
+SAX.elementDecl(doc, 3, ...)
+SAX.externalSubset(doc, , )
+SAX.startElementNs(doc, NULL, NULL, 0, 1, 1, a1='1 2...', 3)
+SAX.endElementNs(doc, NULL, NULL)
+SAX.endDocument()

result/noent/att10

@@ -0,0 +1,18 @@
+<?xml version="1.0"?>
+<!DOCTYPE doc [
+<!ELEMENT doc (tst)*>
+<!ELEMENT tst (#PCDATA)>
+<!ATTLIST tst a NMTOKENS #IMPLIED>
+<!ATTLIST tst b CDATA #IMPLIED>
+<!ENTITY d "&#xD;">
+<!ENTITY a "&#xA;">
+<!ENTITY da "&#xD;&#xA;">
+]>
+<doc>
+<tst a="xyz" b="  xyz"/>
+<tst a="&#13;&#13;A&#10; &#10;B&#13;&#10;" b="&#13;&#13;A&#10; &#10;B&#13;&#10;"/>
+<tst a="&#13;&#13;A&#10;&#10;B&#13;&#10;" b="&#13;&#13;A&#10;&#10;B&#13;&#10;"/>
+<tst a="x y" b=" x  y "/>
+<tst a="a b" b=" a b "/>
+<tst a="a b" b="  a  b  "/>
+</doc>

result/noent/att11

@@ -0,0 +1,9 @@
+<?xml version="1.0" standalone="yes"?>
+<!DOCTYPE attributes [
+<!ELEMENT attributes EMPTY>
+<!ATTLIST attributes nmtoken NMTOKEN #IMPLIED>
+<!ATTLIST attributes nmtokens NMTOKENS #IMPLIED>
+<!ENTITY ent " entity&recursive; ">
+<!ENTITY recursive "reference">
+]>
+<attributes nmtoken="entityreference entityreference entityreference" nmtokens="Test&#13;&#10; this normalization"/>

result/noent/att9

@@ -0,0 +1,6 @@
+<?xml version="1.0"?>
+<!DOCTYPE doc [
+<!ATTLIST doc a1 NMTOKENS "1 2">
+<!ELEMENT doc (#PCDATA)>
+]>
+<doc/>

test/att10

@@ -0,0 +1,22 @@
+<!DOCTYPE doc [
+<!ELEMENT doc (tst*)>
+<!ELEMENT tst (#PCDATA)>
+<!ATTLIST tst a NMTOKENS #IMPLIED>
+<!ATTLIST tst b CDATA #IMPLIED>
+<!ENTITY d "&#xD;">
+<!ENTITY a "&#xA;">
+<!ENTITY da "&#xD;&#xA;">
+]>
+<doc>
+<tst a="
+
+xyz" b="
+
+xyz"/>
+<tst a="&d;&d;A&a;&#x20;&a;B&da;" b="&d;&d;A&a;&#x20;&a;B&da;"/>
+<tst a="&#xd;&#xd;A&#xa;&#xa;B&#xd;&#xa;" b="&#xd;&#xd;A&#xa;&#xa;B&#xd;&#xa;"/>
+<tst a="&#32;x&#32;&#32;y&#32;" b="&#32;x&#32;&#32;y&#32;"/>
+<tst a=" a b " b=" a b "/>
+<tst a="  a  b  " b="  a  b  "/>
+</doc>
+

test/att11

@@ -0,0 +1,13 @@
+<?xml version='1.0' standalone='yes'?>
+<!DOCTYPE attributes [
+<!ELEMENT attributes EMPTY>
+<!ATTLIST attributes
+   nmtoken    NMTOKEN    #IMPLIED
+   nmtokens   NMTOKENS   #IMPLIED>
+<!ENTITY ent " entity&recursive; ">
+<!ENTITY recursive "reference">
+]>
+<attributes
+  nmtoken =  " &ent;   &ent; &ent; "
+  nmtokens = " Test&#x0d;&#x0a;     this&#x20; normalization "
+/>

test/att9

@@ -0,0 +1,5 @@
+<!DOCTYPE doc [
+<!ATTLIST doc a1 NMTOKENS " 1   2       ">
+<!ELEMENT doc (#PCDATA)>
+]>
+<doc/>