#010100: RSS import doesn't work with <dc:> tags

Here is the patch that fixes the problem like you suggested. But for me seems like there is no so much difference when using the example you provided.

Can you please test it and report does it solve this or not?

Index: cronjobs/rssimport.php
===================================================================
--- cronjobs/rssimport.php      (revision 18058)
+++ cronjobs/rssimport.php      (working copy)
@@ -154,7 +154,7 @@

     // Get all items in rss feed
     $itemArray = $root->elementsByName( 'item' );
-    $channel = $root->elementByName( 'channel' );
+    $channel = $root->firstElementByName( 'channel' );

     // Loop through all items in RSS feed
     foreach ( $itemArray as $item )
@@ -183,7 +183,7 @@
     $addCount = 0;

     // Get all items in rss feed
-    $channel =& $root->elementByName( 'channel' );
+    $channel =& $root->firstElementByName( 'channel' );

     // Loop through all items in RSS feed
     foreach ( $channel->elementsByName( 'item' ) as $item )
@@ -225,9 +225,9 @@
     }

     $parentContentObject =& $parentContentObjectTreeNode->attribute( 'object' ); // Get parent content object
-    $titleElement = $item->elementByName( 'title' );
+    $titleElement = $item->firstElementByName( 'title' );
     $title = is_object( $titleElement ) ? $titleElement->textContent() . getCDATA( $titleElement ) : '';
-    $link = $item->elementByName( 'link' );
+    $link = $item->firstElementByName( 'link' );
     $linkURL = $link->textContent() . getCDATA( $link );
     $md5Sum = md5( $linkURL );

@@ -433,7 +433,7 @@
         {
             if ( count( $importDescriptionArray ) == 1 )
             {
-                $element = $xmlDomNode->elementByName( $importDescriptionArray[0] );
+                $element = $xmlDomNode->firstElementByName( $importDescriptionArray[0] );
                 // We should check if text contains CDATA content
                 $resultText = is_object( $element ) ? $element->textContent() . getCDATA( $element ) : false;
                 return $resultText;
@@ -442,7 +442,7 @@
             {
                 $elementName = $importDescriptionArray[0];
                 array_shift( $importDescriptionArray );
-                return recursiveFindRSSElementValue( $importDescriptionArray, $xmlDomNode->elementByName( $elementName ) );
+                return recursiveFindRSSElementValue( $importDescriptionArray, $xmlDomNode->firstElementByName( $elementName ) );
             }
         }

Index: lib/ezxml/classes/ezdomnode.php
===================================================================
--- lib/ezxml/classes/ezdomnode.php     (revision 18057)
+++ lib/ezxml/classes/ezdomnode.php     (working copy)
@@ -363,6 +363,7 @@
     }

     /*!
+      \deprecated Use firstElementByName() instead.
       \returns The first element that is named \a $name.
                If multiple elements with that name is found \c false is returned.

@@ -388,6 +389,27 @@
         return $element;
     }

+    /*
+    \returns The first element that is named \a $name.
+               If multiple elements with that name is found \c false is returned.
+
+      \note This will only make sense for element nodes.
+    */
+    function &firstElementByName( $name )
+    {
+        $element = false;
+        foreach ( array_keys( $this->Children ) as $key )
+        {
+            $child =& $this->Children[$key];
+            if ( $child->name() == $name && !$child->prefix() )
+            {
+                $element =& $child;
+                break;
+            }
+        }
+        return $element;
+    }
+
     /*!
      Alias for libxml compatibility
     */

#251534 by Kirill Subbotin on January 31st, 2007 [Permanent Link]

I added the posted patch, and added the following code to

lib/ezxml/classes/ezdomnode.php:

    /*!
      \returns The first element that is named \a $name.

      \note This will only make sense for element nodes.
      \note Modified version of elementByName, returns first element if multiples are found.
      \note Modification to support RSS
      \sa elementsByName
    */
    function &firstElementByName( $name )
    {
        $element = false;
        foreach ( array_keys( $this->Children ) as $key )
        {
            $child =& $this->Children[$key];
            if ( $child->name() == $name )
            {
                $element =& $child;
                return $element;
            }
        }
        return $element;
    }

The problem remained.

The feed is:

http://content.nejm.org/rss/current.xml

And the node name is blank.

I also tried using <title|alt_title> to name the node, where alt_title was the item pub date. That didn't help either.

eZ 3.8.6

#251550 by Betsy Gamrat on February 2nd, 2007 [Permanent Link]

Sorry, bit slow on that one, got difficulty patching the install.

What version are you using ? trunk 3.9 ?

I'll have to install one that is online with what you have.

X+

PS. Another thing is that I've patched quite a bit already to solve various other RSS bug...

#251566 by Xavier Dutoit on February 5th, 2007 [Permanent Link]

It was trunk (3.10.0) but I think this patch should work on any 3.9 as well.

#251785 by Kirill Subbotin on March 1st, 2007 [Permanent Link]

This problem exists in 4.0.x too

#258522 by Geoff Bentley on October 12th, 2008 [Permanent Link]

Duplicate of #011930: Importing <dc:xxxx> xml fields with RSS

#266753 by Alexandru Stanoi on August 2nd, 2010 [Permanent Link]

I move this from Code Review to Closed since it has been solved in another issue.

Geir Arne Waaler
eZ Documentation

#271048 by Geir Arne Waaler on October 6th, 2011 [Permanent Link]