Enable --revision flag for all deps.

Enable --revision flag for all deps.

Currently, only top-level solutions may be pinned via --revision.  With this, we can:

gclient sync --revision src/third_party/WebKit@deadbeef

BUG=

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=271571

[iannucci, 2014-05-15 18:20:47]

Is that really it? No tests are affected?

[iannucci, 2014-05-15 18:21:12]

On 2014/05/15 18:20:47, iannucci wrote:
> Is that really it? No tests are affected?

(also this will let us eventually ditch custom_vars, woot!)

[M-A Ruel, 2014-05-15 18:22:56]

Interesting idea. It will affect any update step where this warning appeared.
The main downside is that it will silently break, there's no way to know that
sync didn't sync to the expected revision in the case of a typo or a invalid
solution name. Anyhow, I'm fine with this CL. I'll leave approval to Robbie.

[szager1, 2014-05-15 19:05:44]

On 2014/05/15 18:22:56, M-A Ruel wrote:
> Interesting idea. It will affect any update step where this warning appeared.
> The main downside is that it will silently break, there's no way to know that
> sync didn't sync to the expected revision in the case of a typo or a invalid
> solution name. Anyhow, I'm fine with this CL. I'll leave approval to Robbie.

Looks like this breaks --transitive.  Is --transitive actually used anywhere?

[szager1, 2014-05-15 19:06:53]

On 2014/05/15 19:05:44, szager1 wrote:
> On 2014/05/15 18:22:56, M-A Ruel wrote:
> > Interesting idea. It will affect any update step where this warning
appeared.
> > The main downside is that it will silently break, there's no way to know
that
> > sync didn't sync to the expected revision in the case of a typo or a invalid
> > solution name. Anyhow, I'm fine with this CL. I'll leave approval to Robbie.
> 
> Looks like this breaks --transitive.  Is --transitive actually used anywhere?

Actually, I take that back: --transitive works fine.  It's the hack I used to
make gclient_smoketest.py pass that broke --transitive.

*sigh*

New patch forthcoming...

[szager1, 2014-05-15 19:45:10]

OK, unit tests pass, and the error message will be printed any time a --revision
flag does not match any processed dep.

[szager1, 2014-05-19 15:57:56]

ping

[Ryan Tseng, 2014-05-19 22:04:09]

I tried it out and it works, but I have no idea how.  I'll write my nits,
rubberstamp it, and stare at it some more.

https://codereview.chromium.org/290443003/diff/10001/gclient.py
File gclient.py (right):

https://codereview.chromium.org/290443003/diff/10001/gclient.py#newcode606
gclient.py:606: self, command, options, parsed_url, parent):
nit: This fits on the line above, right?

https://codereview.chromium.org/290443003/diff/10001/gclient.py#newcode673
gclient.py:673: revision_overrides.pop(self.name, None)
nit: Move this to line 670, replace revision_overrides.get()?

https://codereview.chromium.org/290443003/diff/10001/gclient.py#newcode1223
gclient.py:1223: sol, rev = revision.split('@', 1)
nit: rename this to (name, rev) to make it make it more clear its not just for
solutions.

[agable, 2014-05-19 22:20:45]

On 2014/05/19 15:57:56, szager1 wrote:
> ping

LGTM. That's... actually pretty elegant. Sweet.

[szager1, 2014-05-20 04:41:33]

https://codereview.chromium.org/290443003/diff/10001/gclient.py
File gclient.py (right):

https://codereview.chromium.org/290443003/diff/10001/gclient.py#newcode606
gclient.py:606: self, command, options, parsed_url, parent):
On 2014/05/19 22:04:09, Ryan T. wrote:
> nit: This fits on the line above, right?

Done.

https://codereview.chromium.org/290443003/diff/10001/gclient.py#newcode673
gclient.py:673: revision_overrides.pop(self.name, None)
On 2014/05/19 22:04:09, Ryan T. wrote:
> nit: Move this to line 670, replace revision_overrides.get()?

Done.

https://codereview.chromium.org/290443003/diff/10001/gclient.py#newcode1223
gclient.py:1223: sol, rev = revision.split('@', 1)
On 2014/05/19 22:04:09, Ryan T. wrote:
> nit: rename this to (name, rev) to make it make it more clear its not just for
> solutions.

Done.

[szager1, 2014-05-20 04:41:43]

The CQ bit was checked by szager@chromium.org

[commit-bot: I haz the power, 2014-05-20 04:41:57]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/szager@chromium.org/290443003/30001

[commit-bot: I haz the power, 2014-05-20 04:43:47]

Change committed as 271571