IDL compiler: switch build to Python compiler

IDL compiler: switch build to Python compiler

This switches the build to Python, rather than Perl.

As of Blink r-167947, Perl and Python generate identical bindings code, so we're ready to switch:
https://src.chromium.org/viewvc/blink?view=rev&revision=167947

Certain in-flight CLs may cause breakage if they land:
if they only change Perl (not Python).
This is generally caught by run-bindings-tests in blink_presubmit, but can fail if the change lacks test cases,
or if the patch set was uploaded prior to turning on Python tests in r-167838.

The only 1 I can see is:
Move mediastream module to oilpan
https://codereview.chromium.org/173363002/
This lacks test cases, and will break the build if it lands.
I've stopped this.

Ref:
log: code_generator_v8.pm
http://src.chromium.org/viewvc/blink/trunk/Source/bindings/scripts/code_generator_v8.pm?view=log#revHEAD

BUG=239771
R=haraken

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=167983

[Nils Barth (inactive), 2014-02-26 04:30:25]

We're ready to switch!

[haraken, 2014-02-26 04:40:03]

OK, LGTM.

https://codereview.chromium.org/166333002/diff/90001/Source/bindings/generate...
File Source/bindings/generated_bindings.gyp (right):

https://codereview.chromium.org/166333002/diff/90001/Source/bindings/generate...
Source/bindings/generated_bindings.gyp:115: 'compiler_module_files': [

idl_compiler_files ?

https://codereview.chromium.org/166333002/diff/90001/Source/bindings/generate...
Source/bindings/generated_bindings.gyp:240:
'<(bindings_output_dir)/JINJA_CACHED',  # Dummy to track dependency

JINJA_CACHED => JinjaTemplate.cached ?

[haraken, 2014-02-26 04:50:53]

Frame is going to be renamed to LocalFrame shortly. You'll need to rebase the CL
after the rename.

[Nils Barth (inactive), 2014-02-26 04:52:09]

On 2014/02/26 04:50:53, haraken wrote:
> Frame is going to be renamed to LocalFrame shortly. You'll need to rebase the
CL
> after the rename.

I'll wait until after the rename before landing, but that shouldn't affect this
CL.

[Nils Barth (inactive), 2014-02-26 06:45:45]

Naming notes, revised.

https://codereview.chromium.org/166333002/diff/90001/Source/bindings/generate...
File Source/bindings/generated_bindings.gyp (right):

https://codereview.chromium.org/166333002/diff/90001/Source/bindings/generate...
Source/bindings/generated_bindings.gyp:115: 'compiler_module_files': [
On 2014/02/26 04:40:04, haraken wrote:
> idl_compiler_files ?

Good point, that's clearer!

https://codereview.chromium.org/166333002/diff/90001/Source/bindings/generate...
Source/bindings/generated_bindings.gyp:240:
'<(bindings_output_dir)/JINJA_CACHED',  # Dummy to track dependency
On 2014/02/26 04:40:04, haraken wrote:
> 
> JINJA_CACHED => JinjaTemplate.cached ?

I'd prefer ALL_CAPS: it's a dummy file (empty),
so it's clearer to not have an extension, as an extension
suggests data type and contents.
(Actual Jinja cache files have a '.cache' extension.)

[haraken, 2014-02-26 06:49:10]

https://codereview.chromium.org/166333002/diff/90001/Source/bindings/generate...
File Source/bindings/generated_bindings.gyp (right):

https://codereview.chromium.org/166333002/diff/90001/Source/bindings/generate...
Source/bindings/generated_bindings.gyp:240:
'<(bindings_output_dir)/JINJA_CACHED',  # Dummy to track dependency
On 2014/02/26 06:45:45, Nils Barth wrote:
> On 2014/02/26 04:40:04, haraken wrote:
> > 
> > JINJA_CACHED => JinjaTemplate.cached ?
> 
> I'd prefer ALL_CAPS: it's a dummy file (empty),
> so it's clearer to not have an extension, as an extension
> suggests data type and contents.
> (Actual Jinja cache files have a '.cache' extension.)

Then JinjaTemplateDummyCache.cached ?

We want to follow the naming convention of Blink unless we have a special
reason.

[Nils Barth (inactive), 2014-02-26 06:52:33]

On 2014/02/26 04:52:09, Nils Barth wrote:
> On 2014/02/26 04:50:53, haraken wrote:
> > Frame is going to be renamed to LocalFrame shortly. You'll need to rebase
the
> CL
> > after the rename.
> 
> I'll wait until after the rename before landing, but that shouldn't affect
this
> CL.

BTW, Frame => LocalFrame did change the compiler (though not this CL),
but Daniel updated Python as well (thanks!), so no need for a separate sync.

[Nils Barth (inactive), 2014-02-26 06:58:53]

On 2014/02/26 06:49:10, haraken wrote:
> We want to follow the naming convention of Blink unless we have a special
> reason.

Oh, you mean we should use CamelCase?
Ok, I've renamed it to: JinjaCachedDummy
(We shouldn't use any extension, because it has no contents.)

[Nils Barth (inactive), 2014-02-26 07:00:46]

Anyway, it's an empty file (not a content-ful file), so it's a bit special.

We use ALLCAPS for special non-source files like DEPS, LICENSE, OWNERS,
PRESUBMIT.py, WATCHLISTS,
which is why I figured it was sensible for a dummy build file.

[abarth-chromium, 2014-02-26 07:02:03]

I think we use ".stamp" files for this purpose in other parts of the build.  I'm
not sure if we have those names in GYP or whether ninja generates them itself.

[Nils Barth (inactive), 2014-02-26 07:11:38]

On 2014/02/26 07:02:03, abarth wrote:
> I think we use ".stamp" files for this purpose in other parts of the build. 
I'm
> not sure if we have those names in GYP or whether ninja generates them itself.

Thanks Adam!
The convention for these "stamp files" is '{target}.stamp';
they are generated by GYP for ninja and make, but don't appear in .gyp files
themselves.
This seems the most standard term, so I've used 'cached_jinja_templates.stamp'
for the file.

Reference:
https://code.google.com/p/gyp/source/browse/trunk/pylib/gyp/generator/make.py...
https://code.google.com/p/gyp/source/browse/trunk/pylib/gyp/generator/ninja.p...

[Nils Barth (inactive), 2014-02-26 07:11:48]

The CQ bit was checked by nbarth@chromium.org

[commit-bot: I haz the power, 2014-02-26 07:11:57]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nbarth@chromium.org/166333002/140001

[Nils Barth (inactive), 2014-02-26 07:27:58]

The CQ bit was unchecked by nbarth@chromium.org

[Nils Barth (inactive), 2014-02-26 11:44:26]

The CQ bit was checked by nbarth@chromium.org

[commit-bot: I haz the power, 2014-02-26 11:44:38]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nbarth@chromium.org/166333002/140001

[Nils Barth (inactive), 2014-02-26 11:44:46]

The CQ bit was unchecked by nbarth@chromium.org

[Nils Barth (inactive), 2014-02-27 01:14:06]

The CQ bit was checked by nbarth@chromium.org

[commit-bot: I haz the power, 2014-02-27 01:14:28]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nbarth@chromium.org/166333002/160001

[commit-bot: I haz the power, 2014-02-27 01:37:26]

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

[commit-bot: I haz the power, 2014-02-27 01:37:27]

Retried try job too often on mac_blink_rel for step(s) blink_heap_unittests,
blink_platform_unittests, webkit_lint, webkit_python_tests, webkit_tests,
webkit_unit_tests, wtf_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_blink_...

[Nils Barth (inactive), 2014-02-27 01:49:03]

The CQ bit was checked by nbarth@chromium.org

[commit-bot: I haz the power, 2014-02-27 01:49:18]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nbarth@chromium.org/166333002/160001

[Nils Barth (inactive), 2014-02-27 01:49:24]

The CQ bit was unchecked by nbarth@chromium.org

[Nils Barth (inactive), 2014-02-27 02:04:14]

On 2014/02/27 01:37:27, I haz the power (commit-bot) wrote:
> Retried try job too often on mac_blink_rel for step(s) blink_heap_unittests,
> blink_platform_unittests, webkit_lint, webkit_python_tests, webkit_tests,
> webkit_unit_tests, wtf_unittests
>
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_blink_...

Real build error, fixed in
IDL compiler: Filter files during pre-caching of templates
https://codereview.chromium.org/182573002/

[Nils Barth (inactive), 2014-02-27 02:04:22]

The CQ bit was checked by nbarth@chromium.org

[commit-bot: I haz the power, 2014-02-27 02:05:32]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nbarth@chromium.org/166333002/180001

[commit-bot: I haz the power, 2014-02-27 04:33:48]

Change committed as 167983

[haraken, 2014-02-27 04:34:39]

On 2014/02/27 04:33:48, I haz the power (commit-bot) wrote:
> Change committed as 167983

How many CLs have you landed until here? :)

[Nils Barth (inactive), 2014-02-27 04:43:48]

On 2014/02/27 04:34:39, haraken wrote:
> On 2014/02/27 04:33:48, I haz the power (commit-bot) wrote:
> > Change committed as 167983
> 
> How many CLs have you landed until here? :)

Probably 400?
...of which I'd estimate about 200 are actual rewrite, 100 are cleanup, and 100
are syncing.

The tracking bug has 400+ comments, most CLs:
Rewrite the IDL compiler in Python
https://code.google.com/p/chromium/issues/detail?id=239771