Attempt at making git-drover work on Windows

Attempt at making git-drover work on Windows

BUG=404755
R=sammc@chromium.org

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=297055

[scottmg, 2015-10-06 18:48:33]

scottmg@chromium.org changed reviewers:
+ sammc@chromium.org

[Peter Kasting, 2015-10-07 22:50:35]

FWIW, I needed to merge a change to M47 today, so I tested this inside msysgit
bash, and it seemed to work correctly.

[iannucci, 2015-10-07 22:57:09]

iannucci@chromium.org changed reviewers:
+ iannucci@chromium.org

[iannucci, 2015-10-07 22:57:09]

lgtm

[iannucci, 2015-10-07 22:58:26]

https://codereview.chromium.org/1387223002/diff/1/git_drover.py
File git_drover.py (right):

https://codereview.chromium.org/1387223002/diff/1/git_drover.py#newcode39
git_drover.py:39: os.symlink = emulate_symlink_windows
Though I'm not totally sold on adding this back into os

Maybe do

if 'nt':
  mk_symlink = shell to mklink
else:
  mk_symlink = os.symlink

?

[scottmg, 2015-10-07 23:03:53]

https://codereview.chromium.org/1387223002/diff/1/git_drover.py
File git_drover.py (right):

https://codereview.chromium.org/1387223002/diff/1/git_drover.py#newcode39
git_drover.py:39: os.symlink = emulate_symlink_windows
On 2015/10/07 22:58:26, iannucci wrote:
> Though I'm not totally sold on adding this back into os
> 
> Maybe do
> 
> if 'nt':
>   mk_symlink = shell to mklink
> else:
>   mk_symlink = os.symlink
> 
> ?

It sort of has to be os.symlink, or else various code in git_common.py code
needs to be changed.

I didn't really digest Primiano's message
https://groups.google.com/a/chromium.org/d/msg/chromium-dev/uVQt89p9ZY0/NZI_6...,
but it sounds like there's a better way than symlinks now anyway? So maybe we
should do this for now and then replace the symlink with alternates later.

[scottmg, 2015-10-07 23:11:33]

The CQ bit was checked by scottmg@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-10-07 23:11:43]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1387223002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1387223002/1

[commit-bot: I haz the power, 2015-10-07 23:13:35]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-10-07 23:13:35]

Dry run: This issue passed the CQ dry run.

[scottmg, 2015-10-07 23:20:40]

I guess it's not so bad. Updated to not monkeypatch 'os'.

[Sam McNally, 2015-10-07 23:29:55]

lgtm

I think this shouldn't work, since some of the symlinks are to files, but msys
seems happy to follow junctions to files so it ends up working.

[scottmg, 2015-10-07 23:30:37]

On 2015/10/07 23:29:55, Sam McNally wrote:
> lgtm
> 
> I think this shouldn't work, since some of the symlinks are to files, but msys
> seems happy to follow junctions to files so it ends up working.

My pseudo-symlink copies if it's a file rather than a dir. Is that what you
mean?

[scottmg, 2015-10-07 23:47:19]

Anyway, I guess I'll land and we can distribute the decision on its workingness
or failingness.

[scottmg, 2015-10-07 23:47:22]

The CQ bit was checked by scottmg@chromium.org

[scottmg, 2015-10-07 23:47:22]

The patchset sent to the CQ was uploaded after l-g-t-m from
iannucci@chromium.org
Link to the patchset: https://codereview.chromium.org/1387223002/#ps20001
(title: "don't monkey patch")

[commit-bot: I haz the power, 2015-10-07 23:47:29]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1387223002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1387223002/20001

[commit-bot: I haz the power, 2015-10-07 23:49:21]

Committed patchset #2 (id:20001) as
http://src.chromium.org/viewvc/chrome?view=rev&revision=297055

[Sam McNally, 2015-10-08 00:12:11]

On 2015/10/07 23:30:37, scottmg wrote:
> On 2015/10/07 23:29:55, Sam McNally wrote:
> > lgtm
> > 
> > I think this shouldn't work, since some of the symlinks are to files, but
msys
> > seems happy to follow junctions to files so it ends up working.
> 
> My pseudo-symlink copies if it's a file rather than a dir. Is that what you
> mean?

Sorry, I misread your change; please disregard.

[mzlan5251_gmail.com, 2015-10-13 05:33:43]

mzlan5251@gmail.com changed reviewers:
+ mzlan5251@gmail.com

[mzlan5251_gmail.com, 2015-10-13 05:33:43]

lgtm

If lny