Convert gclient to use subcommand.py

Convert gclient to use subcommand.py

Update subcommand to support examples, always disable format_description and
format_epilog and add colors when enabled.

R=iannucci@chromium.org
BUG=

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=218180

[M-A Ruel, 2013-08-16 21:08:04]

COLORS!

https://codereview.chromium.org/22824018/diff/1/gclient.py
File gclient.py (right):

https://codereview.chromium.org/22824018/diff/1/gclient.py#newcode7
gclient.py:7: # Files
We don't want that in the gclient help page, that's overwhelming.

https://codereview.chromium.org/22824018/diff/1/gclient.py#newcode1305
gclient.py:1305: Mostly svn-specific. Simply runs 'svn cleanup' for each module.
No need to improperly align these lines anymore.

https://codereview.chromium.org/22824018/diff/1/subcommand.py
File subcommand.py (right):

https://codereview.chromium.org/22824018/diff/1/subcommand.py#newcode178
subcommand.py:178: parser.description = (self.module.__doc__ or '').rstrip()
The following is formatting fine tuning so the output is always well spaced.

[iannucci, 2013-08-16 21:32:05]

Yay colors!

https://codereview.chromium.org/22824018/diff/1/gclient.py
File gclient.py (right):

https://codereview.chromium.org/22824018/diff/1/gclient.py#newcode1305
gclient.py:1305: Mostly svn-specific. Simply runs 'svn cleanup' for each module.
On 2013/08/16 21:08:05, M-A Ruel wrote:
> No need to improperly align these lines anymore.

yay :)

https://codereview.chromium.org/22824018/diff/1/gclient.py#newcode1470
gclient.py:1470: @subcommand.example("""Example:
should the 'Example:' be autogenerated since this api is now called example()?

https://codereview.chromium.org/22824018/diff/1/gclient.py#newcode1528
gclient.py:1528: """Checkout/update all modules.
why 2 lines? Many other docstrings are oneliners

https://codereview.chromium.org/22824018/diff/1/subcommand.py
File subcommand.py (right):

https://codereview.chromium.org/22824018/diff/1/subcommand.py#newcode76
subcommand.py:76: def _is_color_enabled():
_get_color_module() -> colorama or None

https://codereview.chromium.org/22824018/diff/1/subcommand.py#newcode186
subcommand.py:186: lines = (command.__doc__ or '').rstrip().splitlines()
you may want to just use textwrap.dedent()

https://codereview.chromium.org/22824018/diff/1/subcommand.py#newcode201
subcommand.py:201: parser.epilog = getattr(command, 'example', None)
let's call it .example instead, and then add 'Example:' automatically?

[M-A Ruel, 2013-08-17 00:19:09]

https://codereview.chromium.org/22824018/diff/1/gclient.py
File gclient.py (right):

https://codereview.chromium.org/22824018/diff/1/gclient.py#newcode1470
gclient.py:1470: @subcommand.example("""Example:
On 2013/08/16 21:32:05, iannucci wrote:
> should the 'Example:' be autogenerated since this api is now called example()?

Renamed it to epilog so the text can be whatever.

https://codereview.chromium.org/22824018/diff/1/gclient.py#newcode1528
gclient.py:1528: """Checkout/update all modules.
On 2013/08/16 21:32:05, iannucci wrote:
> why 2 lines? Many other docstrings are oneliners

Done. Was the result of improper editing.

https://codereview.chromium.org/22824018/diff/1/subcommand.py
File subcommand.py (right):

https://codereview.chromium.org/22824018/diff/1/subcommand.py#newcode76
subcommand.py:76: def _is_color_enabled():
On 2013/08/16 21:32:05, iannucci wrote:
> _get_color_module() -> colorama or None

Done.

https://codereview.chromium.org/22824018/diff/1/subcommand.py#newcode186
subcommand.py:186: lines = (command.__doc__ or '').rstrip().splitlines()
On 2013/08/16 21:32:05, iannucci wrote:
> you may want to just use textwrap.dedent()

Done.

https://codereview.chromium.org/22824018/diff/1/subcommand.py#newcode201
subcommand.py:201: parser.epilog = getattr(command, 'example', None)
On 2013/08/16 21:32:05, iannucci wrote:
> let's call it .example instead, and then add 'Example:' automatically?

I've decided to switch to naming it epilog, so the users of the library can do
whatever they want.

[M-A Ruel, 2013-08-17 00:40:17]

Bonus change: git cl help status is colorized. I was about to do it in a
separate CL but I had trouble with textwrap, so I merged the CL back and extract
my old dedent() code but didn't fix it. At that point I figure out the error on
my colorizing code but I decided to keep it in this branch since it's relatively
small.

[iannucci, 2013-08-17 01:14:36]

lgtm

https://codereview.chromium.org/22824018/diff/10001/git_cl.py
File git_cl.py (right):

https://codereview.chromium.org/22824018/diff/10001/git_cl.py#newcode1190
git_cl.py:1190: colors = [i for i in dir(Fore) if i[0].isupper()]
Hm... this seems a bit sketchy.

Might as well:

colors = dict((n, v) for n, v in Fore.__dict__.iteritems() if n[0].isupper())
color_re = re.compile("- (%s)" % ("|".join(colors)))

etc.

[M-A Ruel, 2013-08-17 01:22:08]

https://codereview.chromium.org/22824018/diff/10001/git_cl.py
File git_cl.py (right):

https://codereview.chromium.org/22824018/diff/10001/git_cl.py#newcode1190
git_cl.py:1190: colors = [i for i in dir(Fore) if i[0].isupper()]
On 2013/08/17 01:14:36, iannucci wrote:
> Hm... this seems a bit sketchy.
> 
> Might as well:
> 
> colors = dict((n, v) for n, v in Fore.__dict__.iteritems() if n[0].isupper())
> color_re = re.compile("- (%s)" % ("|".join(colors)))
> 
> etc.

If you really prefer but my version is faster. :)

[iannucci, 2013-08-17 01:57:07]

On 2013/08/17 01:22:08, M-A Ruel wrote:
> https://codereview.chromium.org/22824018/diff/10001/git_cl.py
> File git_cl.py (right):
> 
> https://codereview.chromium.org/22824018/diff/10001/git_cl.py#newcode1190
> git_cl.py:1190: colors = [i for i in dir(Fore) if i[0].isupper()]
> On 2013/08/17 01:14:36, iannucci wrote:
> > Hm... this seems a bit sketchy.
> > 
> > Might as well:
> > 
> > colors = dict((n, v) for n, v in Fore.__dict__.iteritems() if
n[0].isupper())
> > color_re = re.compile("- (%s)" % ("|".join(colors)))
> > 
> > etc.
> 
> If you really prefer but my version is faster. :)

faster is better!

[commit-bot: I haz the power, 2013-08-17 15:41:33]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/maruel@chromium.org/22824018/10001

[commit-bot: I haz the power, 2013-08-17 15:43:12]

Presubmit check for 22824018-10001 failed and returned exit status 1.

Running presubmit commit checks ...
Checking out rietveld...
Running save-description-on-failure.sh
Running push-basic.sh
Running upstream.sh
Running submit-from-new-dir.sh
Running abandon.sh
Running submodule-merge-test.sh
Running upload-local-tracking-branch.sh
Running hooks.sh
Running post-dcommit-hook-test.sh
Running upload-stale.sh
Running patch.sh
Running basic.sh

** Presubmit ERRORS **
tests/gclient_smoketest.py failed
....................................F....
======================================================================
FAIL: testSyncJobs (__main__.GClientSmokeSVN)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "tests/gclient_smoketest.py", line 411, in testSyncJobs
    untangle=True)
  File "tests/gclient_smoketest.py", line 84, in parseGclient
    return self.checkBlock(stdout, items)
  File "tests/gclient_smoketest.py", line 144, in checkBlock
    self.checkString(results[i][0][2], path, (i, results[i][0][2], path))
  File
"/b/commit-queue/workdir/tools/depot_tools/testing_support/fake_repos.py", line
739, in checkString
    self.assertEquals(expected, result, msg)
AssertionError: (2, '/tmp/trialjcWbZn/__main__.GClientSmokeSVN.testSyncJobs',
'/tmp/trialjcWbZn/__main__.GClientSmokeSVN.testSyncJobs/src/file/other')

----------------------------------------------------------------------
Ran 41 tests in 41.772s

FAILED (failures=1)


Presubmit checks took 95.7s to calculate.

Was the presubmit check useful? Please send feedback & hate mail to
maruel@chromium.org!

[M-A Ruel, 2013-08-17 16:55:10]

On 2013/08/17 15:43:12, I haz the power (commit-bot) wrote:
> Presubmit check for 22824018-10001 failed and returned exit status 1.
(...)
> FAIL: testSyncJobs (__main__.GClientSmokeSVN)

This was due to non-determinism in two test cases and somehow this code affected
the pseudo-randomness, causing the order of the fetches to be different. I
uploaded patchset 4 that only changes the smoke test to use --jobs 1 instead of
--jobs 8.


> Was the presubmit check useful? Please send feedback & hate mail to
> maruel@chromium.org!

I should change the message here..

[commit-bot: I haz the power, 2013-08-17 16:55:26]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/maruel@chromium.org/22824018/22001

[commit-bot: I haz the power, 2013-08-17 16:57:04]

Change committed as 218180