Bug 117938

Summary: Akregator doesn't follow the Atom spec correctly
Product: [Applications] akregator Reporter: John Maguire <john.maguire>
Component: generalAssignee: kdepim bugs <kdepim-bugs>
Status: RESOLVED FIXED    
Severity: normal    
Priority: NOR    
Version: 1.2   
Target Milestone: ---   
Platform: Gentoo Packages   
OS: Linux   
Latest Commit: Version Fixed In:
Sentry Crash Report:

Description John Maguire 2005-12-08 16:41:24 UTC 
Version:           1.2 (using KDE KDE 3.5.0)
Installed from:    Gentoo Packages
Compiler:          3.3.6 (Gentoo 3.3.6, ssp-3.3.6-1.0, pie-8.7.8) 
OS:                Linux

An example feed that shows the behaviour: http://mostlyharmful.blogdns.com/atom/
If an article is loaded from this atom feed, html tags are visible.

The feed is definitely correct; Akregator has the wrong behaviour as it displays the html markup as text. eg. "<p>" rather than either ignoring it, or parsing the tag.
Atom validator results: http://feedvalidator.org/check.cgi?url=http%3A%2F%2Fmostlyharmful.blogdns.com%2Fatom%2F

In section 4.1.3.3 of the Atom spec (http://www.ietf.org/rfc/rfc4287.txt), it states that all html markup within <content type="html"> should be escaped and that the markup may be used by Atom processors.
Comment 1 Frank Osterfeld 2006-01-16 00:16:41 UTC 
SVN commit 498704 by osterfeld:

fix atom:content parsing: Don't show tags when for Atom 1.0 feeds with escaped HTML in it

BUG: 112491, 117938


 M  +36 -15    tools_p.cpp  


--- branches/KDE/3.5/kdepim/akregator/src/librss/tools_p.cpp #498703:498704
@@ -47,21 +47,42 @@
 	QDomElement e = node.toElement();
 	QString result;
 
-	if (elemName == "content" && ((e.hasAttribute("mode") && e.attribute("mode") == "xml") || !e.hasAttribute("mode")))
-		result = childNodesAsXML(node);
-	else
-		result = e.text();
-
-	bool hasPre = result.contains("<pre>",false);
-	bool hasHtml = hasPre || result.contains("<");	// FIXME: test if we have html, should be more clever -> regexp
-	if(!isInlined && !hasHtml)						// perform nl2br if not a inline elt and it has no html elts
-		result = result = result.replace(QChar('\n'), "<br />");
-	if(!hasPre)										// strip white spaces if no <pre>
-		result = result.simplifyWhiteSpace();
-
-	if (result.isEmpty())
-		return QString::null;
-
+        bool doHTMLCheck = true;
+ 
+        if (elemName == "content") // we have Atom here
+        {
+            doHTMLCheck = false;
+            // the first line is always the Atom 0.3, the second Atom 1.0
+            if (( e.hasAttribute("mode") && e.attribute("mode") == "escaped" && e.attribute("type") == "text/html" )
+            || (!e.hasAttribute("mode") && e.attribute("type") == "html"))
+            {
+                result = KCharsets::resolveEntities(e.text().simplifyWhiteSpace()); // escaped html
+            }
+            else if (( e.hasAttribute("mode") && e.attribute("mode") == "escaped" && e.attribute("type") == "text/plain" )
+                       || (!e.hasAttribute("mode") && e.attribute("type") == "text"))
+            {
+                result = e.text().stripWhiteSpace(); // plain text
+            }
+            else if (( e.hasAttribute("mode") && e.attribute("mode") == "xml" )
+                       || (!e.hasAttribute("mode") && e.attribute("type") == "xhtml"))
+            {
+                result = childNodesAsXML(e); // embedded XHMTL
+            }
+            
+        }
+        
+        if (doHTMLCheck) // check for HTML; not necessary for Atom:content
+        {
+            bool hasPre = result.contains("<pre>",false);
+            bool hasHtml = hasPre || result.contains("<");	// FIXME: test if we have html, should be more clever -> regexp
+            if(!isInlined && !hasHtml)						// perform nl2br if not a inline elt and it has no html elts
+                    result = result = result.replace(QChar('\n'), "<br />");
+            if(!hasPre)										// strip white spaces if no <pre>
+                    result = result.simplifyWhiteSpace();
+        
+            if (result.isEmpty())
+                    return QString::null;
+        }
 	return result;
 }