When processing an atom feed, if the link element follows an image, textinput, summary, tagline, subtitle, logo or info attribute within an entry, that link will be attributed to the feed URL instead of the entry. Atom feeds which do not have a link attribute before those other elements (most, including many atom feed samples around the web), will have the same link URL for every entry as a result.

The aggregator module is virtually unusable for atom feeds unless attributes within the entry are ordered to accommodate.

Parses correctly

<title>This is fine</title>
<link href="http://example.org/" />
<summary>The link appears before the summary.</summary>

Parses incorrectly

<title>We have a problem</title>
<summary>The summary appears before the link, and so we have a problem</summary>
<link href="http://example.org/" />

Comments

carlnewton’s picture

Bump for attention.

carlnewton’s picture

Bump. This is quite a serious issue.

carlnewton’s picture

Bump

carlnewton’s picture

Bump again.

technicalknockout’s picture

I am also seeing this issue. In the modules/aggregator/tests/aggregator_test_atom.xml file, I've messed around with the order of the tags nested in the <entry> tag, and the tests fail.

carlnewton’s picture

Bump

carlnewton’s picture

Still bumping!

technicalknockout’s picture

well, this issue isn't looking like it's getting 'major' priority ... if this is issue is affecting a site you're working on now, you might try using feeds module. I think it's much more robust and flexible. I think it also should be able to do all the core aggregator module does plus more. I haven't used recent versions for aggregation though, so I'd be curious if it works for you.

carlnewton’s picture

Thanks for the heads up. I've asked the developer of a feed we're importing to change the order of attributes to accommodate, so I don't need a patch right now, but I just wanted to make sure this issue got the attention it deserves, but if feeds does exactly what the aggregator module does and more, I suppose that could be why this isn't considered important at the moment.

nsjski’s picture

This is an example of an feed which cannot be imported with aggregator

http://www.birkie.com/news_rss_feed?tags=104564

Aggregator does not find the link or the GUID for the individual feed items. I presume this is part of this issue, because the feed item includes an image before its link and GUID.

<item>
      <title>SAVE! Register by JULY 31 for the Birkie Trail Run Events</title>
      <description>
        <![CDATA[<p><a href='http://www.birkie.com/news_article/show/273489?referral=rss&referrer_id=104650'><img alt='Rails' src='http://cdn3.sportngin.com/attachments/news_article/2632/4720/FINALrsz_btr_randolph_119_2-1_small.jpg' /></a></p>11th Annual RUN, RELAY, and TREK <a href='http://www.birkie.com/news_article/show/273489?referral=rss&referrer_id=104650'>read more...</a><br/><br/>Tag(s): Home]]>
      </description>
      <image>
        <link>http://www.birkie.com/news_article/show/273489?referral=rss&amp;referrer_id=104650</link>
        <title></title>
        <url>http://cdn3.sportngin.com/attachments/news_article/2632/4720/FINALrsz_btr_randolph_119_2-1_small.jpg</url>
      </image>
      <author>Susan Kendrick</author>
      <pubDate>Fri, 26 Jul 2013 15:30:00 -0500</pubDate>
      <link>http://www.birkie.com/news_article/show/273489?referral=rss&amp;referrer_id=104650</link>
      <guid>http://www.birkie.com/news_article/show/273489?referral=rss&amp;referrer_id=104650</guid>
    </item>

I can ask the website owner about it, but their expertise is snow and not software (a nonprofit organization of cross country skiers), so I think my chances of success are limited.

carlnewton’s picture

Yes it seems to be the same issue. I think your options are waiting for a rewrite or using the Feeds module as recommended above. Waiting for a rewrite might take some time because it appears to be a large task, and I'm not sure if anyone is up to it at this point.

nsjski’s picture

Thanks, I'll take a look at Feeds.

nsjski’s picture

Issue summary:View changes

Changed wording

sandykadam’s picture

Issue summary:View changes

@carlnewton

Are you still having issue with this module? because I just checked using different RSS/xml atom feeds and it worked fine for me. Even I tried the URL which gave above.

Please confirm.