Python SCM should be able to retrieve untracked files - Add GIT support

[youenn fablet]

AS illustrated by bug 142400 and 143093, it might be convenient to get untracked files from SCM, either including ignored files or not.

[youenn fablet]

Created attachment 249582 [details]
Patch

[Ryosuke Niwa]

Comment on attachment 249582 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=249582&action=review

> Tools/Scripts/webkitpy/common/checkout/scm/scm_unittest.py:325
> +        # FIXME: Validate that untracked_files do not change with .gitignore/svn:ignore if include_ignored_files is set to True.

Please add a test case for that.  You can, for example, add .pyc file.

[youenn fablet]

Created attachment 249802 [details]
Landing with additional test

[WebKit Commit Bot]

Comment on attachment 249802 [details]
Landing with additional test

Clearing flags on attachment: 249802

Committed r182174: <http://trac.webkit.org/changeset/182174>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[WebKit Commit Bot]

Re-opened since this is blocked by bug 143314

[youenn fablet]

As shown in bug 152257, it would be handy to fix that.
Last time it was landed, it broke windows builds for unobvious reasons.
Is it ok to try relanding it?

[Alex Christensen]

If it broke the Windows build, we should not re-land the same patch, or else it will definitely break the Windows build again.

[youenn fablet]

(In reply to comment #8)
> If it broke the Windows build, we should not re-land the same patch, or else
> it will definitely break the Windows build again.

It happened apparently due to a specific version of cygwin.
Any chance for a change on cygwin usage/cygwin version used on windows bots since April?

[Brent Fulgham]

I believe it broke Windows EWS for reasons I was not able to figure out. I don't think we want to risk breaking EWS again, unless we understand why it broke in the first place.

[youenn fablet]

(In reply to comment #10)
> I believe it broke Windows EWS for reasons I was not able to figure out. I
> don't think we want to risk breaking EWS again, unless we understand why it
> broke in the first place.

IIRC, this caused issues in the parsing of python files.
I was thinking of relanding this patch as several smaller bits: 
- scm generic changes
- svn changes
- git changes, potentially removing the regular expression stuff (root issue maybe?)

[Brent Fulgham]

(In reply to comment #11)
> (In reply to comment #10)
> > I believe it broke Windows EWS for reasons I was not able to figure out. I
> > don't think we want to risk breaking EWS again, unless we understand why it
> > broke in the first place.
> 
> IIRC, this caused issues in the parsing of python files.
> I was thinking of relanding this patch as several smaller bits: 
> - scm generic changes
> - svn changes
> - git changes, potentially removing the regular expression stuff (root issue
> maybe?)

Yes -- I would be in favor of this approach.

[youenn fablet]

Created attachment 268103 [details]
Rebased version

[youenn fablet]

(In reply to comment #13)
> Created attachment 268103 [details]
> Rebased version

Thanks for the review.

The win bots are not yet settled on this patch but this looks unrelated to this patch. And the win bots that tried running this patch seem to continue processing patches.

I plan to land this patch tomorrow.

[youenn fablet]

Created attachment 268361 [details]
Fix bug title in ChangeLog for landing

[WebKit Commit Bot]

Comment on attachment 268361 [details]
Fix bug title in ChangeLog for landing

Clearing flags on attachment: 268361

Committed r194634: <http://trac.webkit.org/changeset/194634>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Brent Fulgham]

All Windows build bots and EWS units are functioning fine with this patch.