Add a flushing mechanism for the WebVTTParser

Add a flushing mechanism for the WebVTTParser

In the WebVTT parser there's an ad-hoc "end-of-stream" check for when
a pending cue should be flushed from the parser. The condition used
however is not reliable given multiple buffers.

To rectify this, add a flush() method to WebVTTParser, and call it
from TextTrackLoader::notifyFinished. Then let this new flush() method
take care of flushing any pending cue from the parser.

BUG=231460
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=161519

[jochen (gone - plz use gerrit), 2013-11-05 10:46:22]

can you add a test for this?

[fs, 2013-11-05 12:33:26]

On 2013/11/05 10:46:22, jochen wrote:
> can you add a test for this?

It would be reasonably easy to do unit test for it - if WebVTTParser did not
have a Document reference. I guess a "http"-test would work as well, haven't
gone through all the resources there, but something like load-and-stall.php
would work. Although, FWIW, this codepath is covered fairly well by the existing
layout tests (any VTT file where the last cue is not followed by an empty line
pretty much). I'll look into doing a http-test, because I'll probably need one
for the next CL of this bug-fix.

[fs, 2013-11-05 14:46:29]

Testcase added. PTAL.

[jochen (gone - plz use gerrit), 2013-11-05 14:50:05]

lgtm, thanks

[commit-bot: I haz the power, 2013-11-05 14:58:51]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fs@opera.com/59453002/60001

[commit-bot: I haz the power, 2013-11-05 17:13:46]

Retried try job too often on win_blink_rel for step(s) webkit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_blink_...

[fs, 2013-11-05 18:01:44]

On 2013/11/05 17:13:46, I haz the power (commit-bot) wrote:
> Retried try job too often on win_blink_rel for step(s) webkit_tests
>
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_blink_...

Looks like the win_blink_rel failed because load-and-stall.php delivered less
than stallAt bytes...Incidently that's what the next CL for this bug should fix.
Any better way to fix this than TestExpectations? Explicit call to ob_start() in
the PHP-script?

[fs, 2013-11-06 07:52:45]

Attempted to prevent EOL normalization on Windows. PTAL again.

[jochen (gone - plz use gerrit), 2013-11-06 08:23:25]

still lgtm

[commit-bot: I haz the power, 2013-11-06 08:32:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fs@opera.com/59453002/210001

[commit-bot: I haz the power, 2013-11-06 10:26:19]

Retried try job too often on win_blink_rel for step(s) webkit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_blink_...

[fs, 2013-11-06 12:18:46]

Made the VTT appear binary, maybe now...

[commit-bot: I haz the power, 2013-11-07 09:11:08]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fs@opera.com/59453002/310002

[commit-bot: I haz the power, 2013-11-07 10:21:00]

Change committed as 161519