395533 - [FIX]No feed preview for http://youtube.com/rss/global/top_favorites.rss

Status: VERIFIED FIXED

(Core :: Networking, defect)

Description

[:Gavin Sharp [email: [e-mail]]]

A recent-ish 1.8 branch build displays a feed preview when visiting http://youtube.com/rss/global/top_favorites.rss . A current trunk build does not. Phil thought this might have been a dupe of bug 364677, but that's a regression in 2.0.0.1 as far as I can tell, while this doesn't seem to have regressed on the branch at all, and the feed does appear to have a <link> in the <channel>.

Comment 1

[:Gavin Sharp [email: [e-mail]]]

Regression range:
http://bonsai.mozilla.org/cvsquery.cgi?module=PhoenixTinderbox&branch=HEAD&branchtype=match&date=explicit&mindate=2007-08-20+18%3A00&maxdate=2007-08-21+06%3A00

It's not bug 385178 (tested by local backout), so I suspect bug 389677 (the feed is served as text/plain). My local backout of that patch is giving me trouble, though, I'll confirm in a bit.

Comment 2

[Phil Ringnalda (:philor)]

Yeah, bug 389677, thus the puzzling way that my copy of the feed at http://dev.philringnalda.com/youtuberss/top_favorites.txt does get preview: it's gzipped, and http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/netwerk/streamconv/converters/nsUnknownDecoder.cpp&rev=1.70&mark=699-700#699 bails on Content-Encoding

Comment 3

[Boris Zbarsky [:bzbarsky]]

Attachment: Proposed fix

Comment 5

[Christian :Biesinger (don't email me, ping me on IRC)]

Hrm, I think it would be better to throw an error instead, as documented in nsIContentSniffer.idl.

Comment 6

[Boris Zbarsky [:bzbarsky]]

Hmm...  Good point.

Comment 7

[Boris Zbarsky [:bzbarsky]]

Attachment: With that change

Comment 8

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 280948 [details] [diff] [review]
With that change

Requesting 1.9 approval for simple regression fix

Comment 9

[Boris Zbarsky [:bzbarsky]]

Checked in.  I'm not sure how to test it.  Does feed preview give a different DOM somehow?  That is, could a mochitest that loads such a URI in a subframe reliably detect that a feed has been sniffed?

Or is this just unit-testable somehow?

Comment 10

[Phil Ringnalda (:philor)]

documentElement.id == "feedHandler" seems like the easiest test for feed preview, offhand.

Comment 11

[Boris Zbarsky [:bzbarsky]]

Is there an obvious DOM that I can try serving as text/plain that would get sniffed as a feed?

Comment 12

[Phil Ringnalda (:philor)]

I was going to attach this, but the minimal feed is actually pretty minimal:

<rss version="2.0">
  <channel>
    <link>http://example.org/</link>
    <title>t</title>
  </channel>
</rss>

(It could be slightly more minimal, with <title/>, but I think that's maybe a bug.) Load it in an iframe, if the iframe's documentElement is an html with id == "feedHandler" then you pass - it displays as plain text in my build from this morning, and gets feed preview in my build from tonight.

Comment 13

[Boris Zbarsky [:bzbarsky]]

Attachment: Mochitest to get landed

Comment 14

[Boris Zbarsky [:bzbarsky]]

Checked in the test.

Comment 15

[Ray Kiddy]

Is it just me? This bug is giving me a failure when I run it locally. I built a copy of FF, checked out 20071202_202630_PST, with mozconfig:
mk_add_options MOZ_CO_PROJECT=browser
ac_add_options --enable-debug
ac_add_options --enable-application=browser
ac_add_options --enable-test
ac_add_options --enable-gtktest
ac_add_options --enable-mochitest
ac_add_options --enable-libIDLtest
ac_add_options --enable-glibtest

I am on Mac OS X 10.4.10 with a fairly standard tool set.

I launched with "perl ./runtests.pl --autorun".

I see this in the web page when I click on the test:

not ok - Text didn't get sniffed as a feed? got "", expected "feedHandler"

and this in the stdout:

--WEBSHELL 0x157e5e80 == 6
++DOMWINDOW == 40
++WEBSHELL 0x2f750b60 == 7
++DOMWINDOW == 41
[loaded plugin /Library/Internet Plug-Ins/QuickTime Plugin.plugin]
++DOMWINDOW == 42
###!!! ASSERTION: cannot call on a dirty frame not currently being reflowed: '!NS_SUBTREE_DIRTY(this) || (GetStateBits() & NS_FRAME_IN_REFLOW)', file nsFrame.cpp, line 556
###!!! ASSERTION: cannot call on a dirty frame not currently being reflowed: '!NS_SUBTREE_DIRTY(this) || (GetStateBits() & NS_FRAME_IN_REFLOW)', file nsFrame.cpp, line 556
###!!! ASSERTION: cannot call on a dirty frame not currently being reflowed: '!NS_SUBTREE_DIRTY(this) || (GetStateBits() & NS_FRAME_IN_REFLOW)', file nsFrame.cpp, line 556
###!!! ASSERTION: cannot call on a dirty frame not currently being reflowed: '!NS_SUBTREE_DIRTY(this) || (GetStateBits() & NS_FRAME_IN_REFLOW)', file nsFrame.cpp, line 556
###!!! ASSERTION: cannot call on a dirty frame not currently being reflowed: '!NS_SUBTREE_DIRTY(this) || (GetStateBits() & NS_FRAME_IN_REFLOW)', file nsFrame.cpp, line 556
###!!! ASSERTION: cannot call on a dirty frame not currently being reflowed: '!NS_SUBTREE_DIRTY(this) || (GetStateBits() & NS_FRAME_IN_REFLOW)', file nsFrame.cpp, line 556
--DOMWINDOW == 41
--DOMWINDOW == 40
--DOMWINDOW == 39
--DOMWINDOW == 38
--DOMWINDOW == 37
--DOMWINDOW == 36

Only one other test is failing. In case it is related, it is:
/tests/content/html/document/test/test_bug391777.html

Comment 16

[Jeff Walden [:Waldo]]

The test was also failing for me a couple weeks back, Fedora 8.  I did a little debugging but didn't get far enough to figure out the problem, and I haven't had time to investigate since then.

Comment 17

[Al Billings [:abillings - ex-MoCo]]

Verified in Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9b3pre) Gecko/2007122704 Minefield/3.0b3pre.

Comment 18

[Olli Pettay [:smaug][[e-mail]]]

For some reason the test fails on my 32 bit Fedora 7, but not on 64 bit Fedora 4 machine.

Comment 19

[Olli Pettay [:smaug][[e-mail]]]

er, 64 bit Fedora 7

Comment 20

[Boris Zbarsky [:bzbarsky]]

Do you see the failure on the site too?  Or just the test?  Does the test obviously fail if you load it standalone?  Want to give debugging it a stab?

Comment 21

[Boris Zbarsky [:bzbarsky]]

Note that we just undid this patch by restricting the types feed sniffing will sniff.