If you write a blog post or whatever kind of post that has a corresponding feed on drupal, and you use a local url, for example <a href="/files/example1.mp3">and then you view this post in Bloglines or FeedBurner, it prepends bloglines domain or feedburner domain onto the url instead of the originating domain. Other readers treat local urls differently and prepend the originating domain, like desktop reader FeedDemon prepends the feed's url, so if the feed url is http://www.developmentseed.org/blog/taxonomy/term/119/0/feed it will make the link to the local file http://www.developmentseed.org/blog/taxonomy/term/119/0/feed/files/examp... where the link should be http://www.developmentseed.org/blog/files/example1.mp3.
Is there a way to fix this in the aggregator to handle each situation? Has anyone else seen this? I looked for an existing issue but saw none, though this seems like a big problem esp. for podcasting. I also tried it out on 4.7 cvs as of Oct. 15.
Ian
| Comment | File | Size | Author |
|---|---|---|---|
| #19 | Screen Shot 2020-10-08 at 4.00.44 pm.png | 90.74 KB | pameeela |
| #19 | Screen Shot 2020-10-08 at 4.02.00 pm.png | 101.62 KB | pameeela |
Comments
Comment #1
beginner commentedNice web site!
Anyway, I confirm the issue (I think we're speaking of the same):
I syndicate www.reuniting.info here:
http://www.masquilier.org/aggregator/sources/2 .
if you look at the August 15, 2005 feed item, there is a local (from the perspective of reuniting.info) link to
resources/questions_and_answers.Of course, it should really be
http://www.reuniting.info/resources/questions_and_answersbut what we get within the aggregator feed page is :http://www.masquilier.org/ resources/questions_and_answerswhich gives plenty of "warning: file not found" in the logs.So, the aggregator module should parse the html, find the base url within the headers of the site, and then add it to the quoted text as appropriate.
Unfortunatly, this parsing stuff is much beyond my abilities for now...
Comment #2
kbahey commentedThis issue http://drupal.org/node/13148 fixes this problem. There is a patch for it.
I will keep this one open though, since it is feed specific. If the other one gets commited, then we can close this one.
Comment #3
beginner commentedHmm... It seems I'm not speaking about the same thing, after all. Note that in my example, the feed is from another site.
Could someone whose blog in another site gets listed here http://drupal.org/aggregator make a test, and create a blog where there would be a relative link to that blog's site. We'll see what happens when the beginning of the blog is displayed at drupal.org...
Comment #4
magico commented@kbahey: what should we do about this? The other issue you refered to was already closed, but I could not find what was opened instead.
Comment #5
lilou commentedComment #6
alex_b commentedThere is a patch for fixing broken relative URLs on aggregation (on the way into Drupal) here #395764: Aggregator: Convert all relative URLs to absolute URLs in feed items .
Is this still a problem on the way out (feed generation)? Moving to node module.
Comment #7
jody lynnYes, I confirmed it's still an issue in /rss.xml.
Comment #9
aspilicious commentedNeeds to be checked in D8
Comment #10
xjm(Merging "node system" and "node.module" components for 8.x; disregard.)
Comment #19
pameeela commentedI can confirm this no longer occurs, see test content with a relative link, which gets the domain added in rss.xml:
