Fixups for libxml bug/feature

git-svn-id: svn+ssh://svn.gna.org/svn/gnustep/libs/base/trunk@16702 72102866-910b-0410-8b05-ffd578937521

ChangeLog

@@ -1,3 +1,11 @@
+2003-05-12  Richard Frith-Macdonald <rfm@gnu.org>
+
+	* Source/Additions/GSXML.m: New -escapedContent method to provide
+	text with the five entities libxml automatically substitutes even
+	when told not to.
+	* Tools/AGSHtml.m: Use -escapedContent to get text suitable for
+	output in html documents.
+
 2003-05-11  Richard Frith-Macdonald <rfm@gnu.org>
 
 	* Tools/gdomap.8.gz: Added man page adapted from internal docs by

Headers/gnustep/base/GSXML.h

@@ -114,6 +114,7 @@
 - (NSDictionary*) attributes;
 - (NSString*) content;
 - (GSXMLDocument*) document;
+- (NSString*) escapedContent;
 - (GSXMLAttribute*) firstAttribute;
 - (GSXMLNode*) firstChild;
 - (GSXMLNode*) firstChildElement;

Source/Additions/GSXML.m

@@ -820,7 +820,13 @@ static NSMapTable	*nodeNames = 0;
  * entity reference nodes, in this case you should not use this
  * method to get the content of the element, but should examine
  * the child nodes of the element individually and perform any
- * entity reference you need to do explicitly.
+ * entity reference you need to do explicitly.<br />
+ * NB. There are five standard entities which are automatically
+ * substituted into the content text rather than appearing as
+ * entity nodes in their own right.  These are '&lt;', '&gt;', '&apos;',
+ * '&quot;' and '&amp;'.  If you with to receive content in which these
+ * characters are represented by the original entity escapes, you need
+ * to use the -escapedContent method.
  */
 - (NSString*) content
 {
@@ -893,6 +899,112 @@ static NSMapTable	*nodeNames = 0;
     }
 }
 
+/**
+ * This performs the same function as the -content method, but retains
+ * the standard five entities (&amp;lt;, &amp;gt;, &amp;apos;, &amp;quot;,
+ * and &amp;amp;) which are normally replaced with their standard equivalents
+ * (&lt;, &gt;, &apos;, &quot;, and &amp;).<br />
+ */
+- (NSString*) escapedContent
+{
+  NSString		*str = [self content];
+
+  if (str != nil)
+    {
+      static NSCharacterSet	*set = nil;
+
+      if (set == nil)
+	{
+	  set = [NSCharacterSet characterSetWithCharactersInString: @"<>'\"&"];
+	  RETAIN(set);
+	}
+      if ([str rangeOfCharacterFromSet: set].length > 0)
+	{
+	  unichar	*base;
+	  unichar	*map;
+	  unichar	c;
+	  unsigned	len;
+	  unsigned	rpos;
+	  unsigned	wpos;
+	  unsigned	end = [str length];
+
+	  base = NSZoneMalloc(NSDefaultMallocZone(), end * sizeof(unichar));
+	  [str getCharacters: base];
+	  for (len = rpos = 0; rpos < end; rpos++)
+	    {
+	      c = base[rpos];
+	      switch (c)
+		{
+		  case '&': 
+		    len += 5;
+		    break;
+		  case '<': 
+		  case '>': 
+		    len += 4;
+		    break;
+		  case '\'': 
+		  case '"': 
+		    len += 6;
+		    break;
+		  default: 
+		    len++;
+		    break;
+		}
+	    }
+	  map = NSZoneMalloc(NSDefaultMallocZone(), len * sizeof(unichar));
+	  for (wpos = rpos = 0; rpos < end; rpos++)
+	    {
+	      c = base[rpos];
+	      switch (c)
+		{
+		  case '&': 
+		    map[wpos++] = '&';
+		    map[wpos++] = 'a';
+		    map[wpos++] = 'm';
+		    map[wpos++] = 'p';
+		    map[wpos++] = ';';
+		    break;
+		  case '<': 
+		    map[wpos++] = '&';
+		    map[wpos++] = 'l';
+		    map[wpos++] = 't';
+		    map[wpos++] = ';';
+		    break;
+		  case '>': 
+		    map[wpos++] = '&';
+		    map[wpos++] = 'g';
+		    map[wpos++] = 't';
+		    map[wpos++] = ';';
+		    break;
+		  case '\'': 
+		    map[wpos++] = '&';
+		    map[wpos++] = 'a';
+		    map[wpos++] = 'p';
+		    map[wpos++] = 'o';
+		    map[wpos++] = 's';
+		    map[wpos++] = ';';
+		    break;
+		  case '"': 
+		    map[wpos++] = '&';
+		    map[wpos++] = 'q';
+		    map[wpos++] = 'u';
+		    map[wpos++] = 'o';
+		    map[wpos++] = 't';
+		    map[wpos++] = ';';
+		    break;
+		  default: 
+		    map[wpos++] = c;
+		    break;
+		}
+	    }
+	  NSZoneFree(NSDefaultMallocZone(), base);
+	  str = [[NSString alloc] initWithCharacters: map length: len];
+	  AUTORELEASE(str);
+	}
+    }
+  return str;
+}
+
 /**
  * Return the first attribute in this node.
  */

Source/GSFormat.m

@@ -922,7 +922,7 @@ NSDictionary *locale)
     if (grouping == (const char *) -1)
       {
 	thousands_sep = [locale objectForKey: NSThousandsSeparator];
-	if (!thousands_sep) thousands_sep = @",";
+	if (thousands_sep == nil) thousands_sep = @",";
 
 	grouping = ""; // FIXME: grouping info missing in locale?
 	if (*grouping == '\0' || *grouping == CHAR_MAX)
@@ -1405,6 +1405,7 @@ NSDictionary *locale)
 	NSString	*decimal_sep;
 
 	decimal_sep = [locale objectForKey: NSDecimalSeparator];
+	if (decimal_sep == nil) decimal_sep = @".";
 
 	bp = buf1;
 
@@ -1525,6 +1526,7 @@ NSDictionary *locale)
 	NSString	*decimal_sep;
 
 	decimal_sep = [locale objectForKey: NSDecimalSeparator];
+	if (decimal_sep == nil) decimal_sep = @".";
 
 	bp = buf1;
 

Tools/AGSHtml.m

@@ -774,7 +774,7 @@ static NSMutableSet	*textNodes = nil;
 	  NSString		*text;
 	  NSString		*val;
 
-	  text = [children content];
+	  text = [children escapedContent];
 	  val = [prop objectForKey: @"id"];
 	  if (val == nil)
 	    {
@@ -847,7 +847,7 @@ static NSMutableSet	*textNodes = nil;
 			{
 			  if ([t type] == XML_TEXT_NODE)
 			    {
-			      NSString	*content = [t content];
+			      NSString	*content = [t escapedContent];
 
 			      if (content == nil) content = @"";
 			      str = [str stringByAppendingString: content];
@@ -1182,7 +1182,7 @@ static NSMutableSet	*textNodes = nil;
 	  NSString		*text;
 	  NSString		*val;
 
-	  text = [children content];
+	  text = [children escapedContent];
 	  val = [prop objectForKey: @"id"];
 	  if (val == nil)
 	    {
@@ -1225,7 +1225,7 @@ static NSMutableSet	*textNodes = nil;
 			{
 			  if ([t type] == XML_TEXT_NODE)
 			    {
-			      NSString	*content = [t content];
+			      NSString	*content = [t escapedContent];
 
 			      if (content == nil) content = @"";
 			      str = [str stringByAppendingString: content];
@@ -1340,7 +1340,7 @@ static NSMutableSet	*textNodes = nil;
 			{
 			  if ([t type] == XML_TEXT_NODE)
 			    {
-			      NSString	*content = [t content];
+			      NSString	*content = [t escapedContent];
 
 			      if (content == nil) content = @"";
 			      sel = [sel stringByAppendingString: content];
@@ -1366,7 +1366,7 @@ static NSMutableSet	*textNodes = nil;
 			{
 			  if ([t type] == XML_TEXT_NODE)
 			    {
-			      NSString	*content = [t content];
+			      NSString	*content = [t escapedContent];
 			
 			      if (content == nil) content = @"";
 			      str = [str stringByAppendingString: content];
@@ -2002,7 +2002,7 @@ NSLog(@"Element '%@' not implemented", name); // FIXME
     {
       if ([node type] == XML_TEXT_NODE)
 	{
-	  NSString	*str = [node content];
+	  NSString	*str = [node escapedContent];
 
 	  if (str == nil) str = @"";
 	  [buf appendString: str];
@@ -2055,7 +2055,7 @@ NSLog(@"Element '%@' not implemented", name); // FIXME
       [buf appendString: @"<dt><b>Conforms to:</b></dt>\n"];
       while (node != nil && [[node name] isEqual: @"conform"] == YES)
 	{
-	  NSString	*text = [[node firstChild] content];
+	  NSString	*text = [[node firstChild] escapedContent];
 
 	  if (text == nil) text = @"";
 	  [buf appendString: indent];

Tools/autogsdoc.m

@@ -1196,9 +1196,9 @@ main(int argc, char **argv, char **env)
 		  AGSIndex	*localRefs;
 
 		  parser = [GSXMLParser parserWithContentsOfFile: gsdocfile];
-		  [parser substituteEntities: NO];
 		  [parser doValidityChecking: YES];
 		  [parser keepBlanks: NO];
+		  [parser substituteEntities: NO];
 		  if ([parser parse] == NO)
 		    {
 		      NSLog(@"WARNING %@ is not a valid document", gsdocfile);
@@ -1485,9 +1485,9 @@ main(int argc, char **argv, char **env)
 			file, gDate, hDate);
 		    }
 		  parser = [GSXMLParser parserWithContentsOfFile: gsdocfile];
-		  [parser substituteEntities: NO];
 		  [parser doValidityChecking: YES];
 		  [parser keepBlanks: NO];
+		  [parser substituteEntities: NO];
 		  if ([parser parse] == NO)
 		    {
 		      NSLog(@"WARNING %@ is not a valid document", gsdocfile);