Bug 39372

Summary:

check-webkit-style complains about use of NULL in GTK function calls that require sentinels

Product:

WebKit

Reporter:

Darin Adler <darin>

Component:

Tools / Tests

Assignee:

Nobody <webkit-unassigned>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

abarth, cjerdonek, eric, hamaji, levin, pnormand

Priority:

Version:

528+ (Nightly build)

Hardware:

All

OS:

All

Attachments:

Description	Flags
proposed patch	hamaji: review+

Darin Adler

Reported 2010-05-19 11:57:42 PDT

Various GTK calls have variable arguments that are pointers, and then have to end with a sentinel. The sentinel must be NULL, not 0. But check-webkit-style complains, saying that the NULL should instead be 0.

Attachments
proposed patch (2.51 KB, patch) 2010-05-27 02:33 PDT, Philippe Normand	hamaji: review+	Details Formatted Diff Diff
View All Add attachment proposed patch, testcase, etc.

Darin Adler

Comment 1 2010-05-19 11:58:04 PDT

See an example in bug 39368.

Philippe Normand

Comment 2 2010-05-27 02:33:05 PDT

Created attachment 57212 [details] proposed patch

Shinichiro Hamaji

Comment 3 2010-05-27 03:01:18 PDT

Comment on attachment 57212 [details] proposed patch Looks good, but it would be better if we can limit the symbols. How about r'\bg_[a-z]+_[a-z]+\b', r'\bg_[a-z]+(_[a-z]+)+\b', or r'\bg(_[a-z]+)+\b' ?

Philippe Normand

Comment 4 2010-05-27 05:10:30 PDT

(In reply to comment #3) > (From update of attachment 57212 [details]) > Looks good, but it would be better if we can limit the symbols. How about r'\bg_[a-z]+_[a-z]+\b', r'\bg_[a-z]+(_[a-z]+)+\b', or r'\bg(_[a-z]+)+\b' ? Yeah, I like option #3. Will update the patch and land it, thanks for the review :)

Philippe Normand

Comment 5 2010-05-27 05:15:39 PDT

Landed in r60293 <http://trac.webkit.org/changeset/60293> Thanks!

Note You need to log in before you can comment on or make changes to this bug.