Bug 43766 - Clients of HTMLTokenizer should be able to see where characters went in the token
Summary: Clients of HTMLTokenizer should be able to see where characters went in the t...
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: New Bugs (show other bugs)
Version: 528+ (Nightly build)
Hardware: Other OS X 10.5
: P2 Normal
Assignee: Adam Barth
URL:
Keywords:
Depends on:
Blocks: 43746
  Show dependency tree / graph
 
Reported: 2010-08-09 20:07 PDT by Adam Barth
Modified: 2010-08-10 18:56 PDT (History)
5 users (show)

See Also:

Attachments
Patch (11.15 KB, patch)
2010-08-09 20:11 PDT, Adam Barth 		no flags Details | Formatted Diff | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Adam Barth 2010-08-09 20:07:54 PDT 
Clients of HTMLTokenizer should be able to see where characters when in the token
Comment 1 Adam Barth 2010-08-09 20:11:16 PDT 
Created attachment 63969 [details]
Patch
Comment 2 Adam Barth 2010-08-09 20:11:42 PDT 
This facility will be used to implement view-source in Bug 43746.
Comment 3 Early Warning System Bot 2010-08-09 21:04:59 PDT 
Attachment 63969 [details] did not build on qt:
Build output: http://queues.webkit.org/results/3705014
Comment 4 Eric Seidel (no email) 2010-08-09 23:15:39 PDT 
Attachment 63969 [details] did not build on mac:
Build output: http://queues.webkit.org/results/3772008
Comment 5 WebKit Review Bot 2010-08-09 23:37:15 PDT 
Attachment 63969 [details] did not build on chromium:
Build output: http://queues.webkit.org/results/3728008
Comment 6 Eric Seidel (no email) 2010-08-10 00:37:34 PDT 
Comment on attachment 63969 [details]
Patch

I worry there is an off-by-one typo waiting to happen here.

Looks OK AFAICT.

Did you test perf?
Comment 7 Adam Barth 2010-08-10 12:03:19 PDT 
Comment on attachment 63969 [details]
Patch

> Did you test perf?

I did.  It showed a minor slowdown (~1%) on the parser benchmark.  We discussed in person whether we should worry about these things.  I'm inclined to land this now (to unblock the view-source + removing the old parser work), but I'll play with it a bit to understand the performance sensitivity.  I suspect it comes from adding 4 ints to the HTMLToken::Attribute structure, but we'll find out.
Comment 8 WebKit Commit Bot 2010-08-10 18:56:18 PDT 
Comment on attachment 63969 [details]
Patch

Clearing flags on attachment: 63969

Committed r65110: <http://trac.webkit.org/changeset/65110>
Comment 9 WebKit Commit Bot 2010-08-10 18:56:24 PDT 
All reviewed patches have been landed.  Closing bug.