Make PRESUBMIT.py run checks via 'pcg' if it is available in PATH.

Make PRESUBMIT.py run checks via 'pcg' if it is available in PATH.

R=maruel@chromium.org, iannucci@chromium.org
BUG=chromium:529936
Committed: https://github.com/luci/gae/commit/ca8b17b329e830921994fcdb8d3fc79cbe02d144

[Vadim Sh., 2015-09-14 00:29:31]

PTAL

Example of error:

$ git cl presubmit --force
Running presubmit commit checks ...

** Presubmit ERRORS **
pre-commit-go: lint, pre-commit, pre-push (17.54s) failed
these files are improperly formmatted, please run: goimports -w <files>
service/info/context.go

golint failed:
service/info/context.go:9:2: a blank import should be only in a main or test
package, or have a comment justifying it
pcg: checks failed in 17.49s


Presubmit checks took 17.6s to calculate.

[iannucci, 2015-09-14 00:41:11]

I'll have to take a look at this to fix the warnings. What do you think about
putting pre commit go in presubmit canned checks?

[Vadim Sh., 2015-09-14 00:43:15]

On 2015/09/14 00:41:11, iannucci wrote:
> I'll have to take a look at this to fix the warnings.
It checks only diffs. So if all new code is warning-free, it should be fine to
commit now. The error in example was intentionally induced. 

>  What do you think about putting pre commit go in presubmit canned checks?
Maybe later when we use it for awhile and know how it behaves.

[M-A Ruel, 2015-09-14 01:37:33]

On 2015/09/14 00:43:15, Vadim Sh. wrote:
> On 2015/09/14 00:41:11, iannucci wrote:
> > I'll have to take a look at this to fix the warnings.
> It checks only diffs. So if all new code is warning-free, it should be fine to
> commit now. The error in example was intentionally induced. 

Sadly, this is not 100% true for some checks. I forget which but some are
basically running on all the code all the time. :/

Also, having a pure lint-free code is not necessarily a great idea IMHO.


> >  What do you think about putting pre commit go in presubmit canned checks?
> Maybe later when we use it for awhile and know how it behaves.

Agreed.

[Vadim Sh., 2015-09-14 17:55:36]

On 2015/09/14 01:37:33, M-A Ruel wrote:
> On 2015/09/14 00:43:15, Vadim Sh. wrote:
> > On 2015/09/14 00:41:11, iannucci wrote:
> > > I'll have to take a look at this to fix the warnings.
> > It checks only diffs. So if all new code is warning-free, it should be fine
to
> > commit now. The error in example was intentionally induced. 
> 
> Sadly, this is not 100% true for some checks. I forget which but some are
> basically running on all the code all the time. :/
> 
> Also, having a pure lint-free code is not necessarily a great idea IMHO.
> 
infra.git/go is passing golint and go vet and it wasn't a big problem to keep it
that way.
I'm not sure about errcheck though. 

In general I think if linting errors are not forcefully checked they are ignored
:) Let's try to 
run linters on changes and see how it goes. IIUC, pcg supports blacklists for
some checks,
perhaps it will be enough.

> 
> > >  What do you think about putting pre commit go in presubmit canned checks?
> > Maybe later when we use it for awhile and know how it behaves.
> 
> Agreed.

[M-A Ruel, 2015-09-14 20:14:11]

lgtm

https://codereview.chromium.org/1342673002/diff/1/PRESUBMIT.py
File PRESUBMIT.py (right):

https://codereview.chromium.org/1342673002/diff/1/PRESUBMIT.py#newcode54
PRESUBMIT.py:54: r'.*? Copyright %(year)s The Chromium Authors\. '
I think we should standardize to LUCI Authors ASAP.

[Vadim Sh., 2015-09-14 20:35:45]

The CQ bit was checked by vadimsh@chromium.org

[commit-bot: I haz the power, 2015-09-14 20:36:02]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1342673002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1342673002/1

[commit-bot: I haz the power, 2015-09-14 20:38:29]

Committed patchset #1 (id:1) as
https://github.com/luci/gae/commit/ca8b17b329e830921994fcdb8d3fc79cbe02d144