#147 ✓resolved
Kieran P

Image RSS feed broken, and output not escaped

Reported by Kieran P | January 5th, 2009 @ 02:02 PM | in 1.2

The image rss feeds are broken because of a nil.public_filename errors. It looks like the image partial for the rss feed is being passed the object with the wrong local name the partial is expecting.

Also, the images don't show up in the rss feeds because the content of in is not escaped or decoded. Suggest adding htmlentities require, decoding entities, and then escaping the output like we do for item rss feed descriptions.

Comments and changes to this ticket

  • Walter McGinnis

    Walter McGinnis January 5th, 2009 @ 02:44 PM

    • Tag changed from bug, error, partial, rss to bug, error, mrss, pagination, partial, rss

    First up is to add tests for each existing rss feed please.

    Then, while we are in there fixing this issue (but we should still open a specific ticket for the enhancement), we should add support for CoolIris (http://www.cooliris.com/) which means two things:

    Cheers, Walter

  • Walter McGinnis

    Walter McGinnis January 5th, 2009 @ 02:47 PM

    The tests should also validate the returned xml. That is one thing that keeps cropping up (probably related to this escaping issue).

  • Kieran P

    Kieran P January 5th, 2009 @ 03:51 PM

    • State changed from “new” to “to-review”

    The fix for the rss feeds has been applied and pushed to branch

    bugfix_lh147_image_rss_feed_broken

    http://github.com/kete/kete/comm...

    I'll open a new ticket for the CoolIris work.

  • Kieran P

    Kieran P January 5th, 2009 @ 05:25 PM

    I've pushed fixes which allow for non alphanumeric searches and fixing other various issues.

    http://github.com/kete/kete/comm...

    http://github.com/kete/kete/comm...

    Basically Zebra now takes raw text, escapes it and stores.

    Searching now takes raw text, escapes it and searches.

    So the display and searching with non alphanumeric chars is more resilient now.

    If these changes are alright, it might be worth continuing and trying to get ampersands working (the same code now should just work hopefully).

  • Walter McGinnis

    Walter McGinnis January 5th, 2009 @ 06:34 PM

    Yep, I agree about ampersands. At the least tags, titles, and creator/contributors in lib/oai_dc_helpers.rb will need escaping. You'll want to run through that file and see if you catch anything else that might need escaping (extended fields with xml elment name that dc schema values?).

    I think we need to be pretty conservative with this move, so what I recommend is we limit the commits here that we merge immediately to master (rss fixes/tests only) and cherry pick the others into a new branch based on #52 and #148, i.e. bugfixes-52-148-combined or something). Then we'll thoroughly review and test that branch for unexpected results.

  • Walter McGinnis

    Walter McGinnis January 6th, 2009 @ 04:25 PM

    • State changed from “to-review” to “open”

    I took a look at http://github.com/kete/kete/tree... and it seems pretty good. The only concern I have is that it has the commits that have been moved to #52, etc and we don't want these moved into master yet. They will be handled by their dedicated branch.

    If it does have the stuff from the other branch, please cherry pick around them into master. If not, just merge this branch to master.

  • Kieran P

    Kieran P January 6th, 2009 @ 07:37 PM

    • State changed from “open” to “resolved”

    I split the two branches as mentioned above. http://github.com/kete/kete/tree... doesn't have any of the search related fixes, so I'll merge that into master now.

    Resolving ticket (I'll open up another for the CoolIris work when it's started).

Please Sign in or create a free account to add a new ticket.

With your very own profile, you can contribute to projects, track your activity, watch tickets, receive and update tickets through your email and much more.

New-ticket Create new ticket

Create your profile

Help contribute to this project by taking a few moments to create your personal profile. Create your profile ยป

Kete was developed by Horowhenua Library Trust and Katipo Communications Ltd. to build a digital library of Horowhenua material.

People watching this ticket

Pages