Support -Goutput_dir=blahblah in GN-GYP hybrid mode

Support -Goutput_dir=blahblah in GN-GYP hybrid mode

This supports specifying the output_dir generator flag in the gn-generating-gyp
mode. The value of the output_dir flag is mapped to a gyp variable called gyp_output_dir
and a GN argument by the same name. References from gyp to gn-generated gyp files must
use this variable and references from BUILD.gn files must use the gn argument.

BUG=335760
TBR=brettw for build/config/BUILDCONFIG.gn (looked at in person)

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=245800

[Nico, 2014-01-18 00:38:44]

lgtm

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium
File build/gyp_chromium (right):

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
build/gyp_chromium:144: return sys.argv[i][13:]
s/13/len('-Goutput_dir=')/

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
build/gyp_chromium:146: return "out_foo"
_foo?

[jamesr, 2014-01-18 00:39:52]

On 2014/01/18 00:38:44, Nico wrote:
> lgtm
> 
> https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium
> File build/gyp_chromium (right):
> 
>
https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
> build/gyp_chromium:144: return sys.argv[i][13:]
> s/13/len('-Goutput_dir=')/
> 
>
https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
> build/gyp_chromium:146: return "out_foo"
> _foo?

whoops, some local testing that snuck through

[piman, 2014-01-18 00:44:27]

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium
File build/gyp_chromium (right):

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
build/gyp_chromium:133: for i in range(len(env_items)):
nit: 'for item in env_items' is simpler

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
build/gyp_chromium:134: print env_items[i]
hi

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
build/gyp_chromium:136: print "yay %s %s" % (env_items[i], env_items[i][12:])
yay!

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
build/gyp_chromium:142: for i in range(len(sys.argv))[1:]:
for arg in sys.argv[1:]

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
build/gyp_chromium:144: return sys.argv[i][13:]
arg.split("=")[2] ?

[Nico, 2014-01-18 00:46:52]

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium
File build/gyp_chromium (right):

https://codereview.chromium.org/142223002/diff/40001/build/gyp_chromium#newco...
build/gyp_chromium:144: return sys.argv[i][13:]
On 2014/01/18 00:44:27, piman wrote:
> arg.split("=")[2] ?

arg.split("=")[1] ?

[jamesr, 2014-01-18 00:48:20]

Ah you split people.  What if someone wants -"Goutput_dir=my=favorite=directory"
?

It's now startswith(needle) / foo[len(needle):]

[Nico, 2014-01-18 00:50:23]

'a=b=c'.split('=', 1)[1]


On Fri, Jan 17, 2014 at 4:48 PM, <jamesr@chromium.org> wrote:

> Ah you split people.  What if someone wants -"Goutput_dir=my=favorite=
> directory"
> ?
>
> It's now startswith(needle) / foo[len(needle):]
>
> https://codereview.chromium.org/142223002/
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[piman, 2014-01-18 00:54:51]

https://codereview.chromium.org/142223002/diff/120001/build/gyp_chromium
File build/gyp_chromium (right):

https://codereview.chromium.org/142223002/diff/120001/build/gyp_chromium#newc...
build/gyp_chromium:136: return item[len(needle):]
BTW, does the env variable override the flag, or the opposite (it's usually the
opposite)?

[Nico, 2014-01-18 00:59:30]

On Fri, Jan 17, 2014 at 4:54 PM, <piman@chromium.org> wrote:

>
> https://codereview.chromium.org/142223002/diff/120001/build/gyp_chromium
> File build/gyp_chromium (right):
>
> https://codereview.chromium.org/142223002/diff/120001/
> build/gyp_chromium#newcode136
> build/gyp_chromium:136: return item[len(needle):]
> BTW, does the env variable override the flag, or the opposite (it's
> usually the opposite)?
>

From tools/gyp/pylib/gyp/__init__.py:


  if not options.generator_output and options.use_environment:
    g_o = os.environ.get('GYP_GENERATOR_OUTPUT')
    if g_o:
      options.generator_output = g_o

So the env var is only used if no -G flag is present
…apparently gyp also has a --ignore-environment flag which disables

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[jamesr, 2014-01-18 01:00:24]

On 2014/01/18 00:54:51, piman wrote:
> https://codereview.chromium.org/142223002/diff/120001/build/gyp_chromium
> File build/gyp_chromium (right):
> 
>
https://codereview.chromium.org/142223002/diff/120001/build/gyp_chromium#newc...
> build/gyp_chromium:136: return item[len(needle):]
> BTW, does the env variable override the flag, or the opposite (it's usually
the
> opposite)?

Ah you're right - I thought it was the other way around but testing shows that
the gyp code's behavior is that the flag overrides the env variable. Will flip
this function.

[Nico, 2014-01-18 01:00:51]

Sorry, gmail sent this in the middle of me typing it.

From ..._init__.py:

  gen_flags = []
  if options.use_environment:
    gen_flags += ShlexEnv('GYP_GENERATOR_FLAGS')
  if options.generator_flags:
    gen_flags += options.generator_flags
  generator_flags = NameValueListToDict(gen_flags)

So both are set appended, and I think the latter wins in
NameValueListToDict.

(--ingore-environment is ignored for GYP_GENERATOR_FLAGS)


On Fri, Jan 17, 2014 at 4:59 PM, Nico Weber <thakis@chromium.org> wrote:

> On Fri, Jan 17, 2014 at 4:54 PM, <piman@chromium.org> wrote:
>
>>
>> https://codereview.chromium.org/142223002/diff/120001/build/gyp_chromium
>> File build/gyp_chromium (right):
>>
>> https://codereview.chromium.org/142223002/diff/120001/
>> build/gyp_chromium#newcode136
>> build/gyp_chromium:136: return item[len(needle):]
>> BTW, does the env variable override the flag, or the opposite (it's
>> usually the opposite)?
>>
>
> From tools/gyp/pylib/gyp/__init__.py:
>
>
>   if not options.generator_output and options.use_environment:
>     g_o = os.environ.get('GYP_GENERATOR_OUTPUT')
>     if g_o:
>       options.generator_output = g_o
>
> So the env var is only used if no -G flag is present
> …apparently gyp also has a --ignore-environment flag which disables
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[piman, 2014-01-18 01:21:24]

lgtm

[commit-bot: I haz the power, 2014-01-18 01:26:11]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/142223002/190001

[jamesr, 2014-01-18 02:48:21]

Brett - doesn't look like this will make it through the CQ.  Feel free to take
over, or I can try to land it on top of whatever's in the tree tomorrow.  I
think I'll land it in two parts - first the gyp_chromium/BUILDCONFIG.gn changes
to define and use the new variable in gn-generated code and then the changes to
all of the .gyp files to refer to the new location.  That way if we need to
revert things we can revert the changes to the non-generated .gyp files without
having to back the rest out.

[commit-bot: I haz the power, 2014-01-18 22:03:57]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/142223002/190001

[commit-bot: I haz the power, 2014-01-18 23:51:44]

Change committed as 245800