[Extensions] Add extension feature generation code

[Extensions] Add extension feature generation code

Extension features files are json files that contain information about
which APIs or extension features can be used in which contexts.
Currently, these files are parsed at runtime during both browser
startup and renderer process startup, causing a significant delay.

This patch implements the bulk of the functionality to move towards
generating feature files rather than parsing them from the json, which
should hopefully result in a significant speed improvement.

This introduces the feature_compiler.py, which generates the .h and .cc
files for a given set of features.json files, as well as the build rules
to invoke it automatically and tests for it. This does not add generated
files to the production build of chrome yet.

Sample api_feature_provider.h and .cc:
https://gist.github.com/anonymous/9bb094a90810b6050556761c87c204ac

BUG=280286
Committed: https://crrev.com/88155eb86d2b8b8493eb80d09feb639d34dfdcc4
Cr-Commit-Position: refs/heads/master@{#406575}

[Devlin, 2016-07-14 00:27:25]

Description was changed from

==========
Generate!

BUG=
==========

to

==========
[Extensions] Add extension feature generation code

Extension features files are json files that contain information about
which APIs or extension features can be used in which contexts.
Currently, these files are parsed at runtime during both browser
startup and renderer process startup, causing a significant delay.

This patch implements the bulk of the functionality to move towards
generating feature files rather than parsing them from the json, which
should hopefully result in a significant speed improvement.

This introduces the feature_compiler.py, which generates the .h and .cc
files for a given set of features.json files, as well as the build rules
to invoke it automatically and tests for it. This does not add generated
files to the production build of chrome yet.

Sample api_feature_provider.h and .cc:
https://gist.github.com/anonymous/9bb094a90810b6050556761c87c204ac

BUG=280286
==========

[Devlin, 2016-07-14 00:27:36]

Patchset #1 (id:1) has been deleted

[Devlin, 2016-07-14 00:27:41]

Patchset #1 (id:20001) has been deleted

[Devlin, 2016-07-14 00:28:44]

rdevlin.cronin@chromium.org changed reviewers:
+ asargent@chromium.org

[Devlin, 2016-07-14 00:28:45]

Antony, wanna take a look?  I still need to add the gyp versions of the compile,
but I think this is getting close to ready.

Let me know if you have any questions!

[Devlin, 2016-07-14 22:23:20]

Patchset #3 (id:80001) has been deleted

[Devlin, 2016-07-14 22:27:20]

rdevlin.cronin@chromium.org changed reviewers:
+ dpranke@chromium.org

[Devlin, 2016-07-14 22:27:21]

+Dirk.

Dirk, some quick context here - I'm working on moving our (extensions) feature
files from being json that's parsed at runtime during startup (yuck!) to
generated C++.

Unfortunately, neither Antony nor myself are really familiar with gyp/gn.  Do
you mind taking a look at the gyp/gn files - in particular, the new actions in
//build?  I'm sure right now there's a lot of bizarre stuff I'm doing as a
result of the copy-paste-hack approach I took.

[Dirk Pranke, 2016-07-14 23:16:36]

This basically looks okay, just some stylistic stuff ...

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gni
File build/json_features.gni (right):

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:1: # Copyright 2016 The Chromium Authors. All rights
reserved.
Where is this template actually going to be used?

We generally don't want to put chromium-specific stuff in //build, as we want to
get //build to be a generic things that can be reused by other (non-chromium)
projects.

Can you move these to //tools/json_schema_compiler/json_features.gni or
something like that?

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:27: script = compiler_script
I would get rid of compiler script and just refer to
"$compiler_root/feature_compiler.py" directly.

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:47: public_deps = lib_public_deps
similarly, just inline the values here.

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gypi
File build/json_features.gypi (right):

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gy...
build/json_features.gypi:31: '<(compiler_dir)/feature_compiler.py',
'<(compiler_script)' ?

[Devlin, 2016-07-15 16:19:07]

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gni
File build/json_features.gni (right):

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:1: # Copyright 2016 The Chromium Authors. All rights
reserved.
On 2016/07/14 23:16:36, Dirk Pranke wrote:
> Where is this template actually going to be used?
> 
> We generally don't want to put chromium-specific stuff in //build, as we want
to
> get //build to be a generic things that can be reused by other (non-chromium)
> projects.
> 
> Can you move these to //tools/json_schema_compiler/json_features.gni or
> something like that?

This is going to be used from //chrome and //extensions.

I just copied where the other similar templates were. :) We can move this, as
long as there's no qualms about including tools/json_schema_compiler templates
from chromium sources.

Done.

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:27: script = compiler_script
On 2016/07/14 23:16:35, Dirk Pranke wrote:
> I would get rid of compiler script and just refer to
> "$compiler_root/feature_compiler.py" directly.

Done.

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:47: public_deps = lib_public_deps
On 2016/07/14 23:16:36, Dirk Pranke wrote:
> similarly, just inline the values here.

This seems to work for public deps, but trying to inline sources gives an error:
ERROR at //build/json_features.gni:43:33: Target not found in this context.
    sources = invoker.sources + get_target_outputs(":target")
                                ^---------------------------

(Done for public_deps)

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gypi
File build/json_features.gypi (right):

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gy...
build/json_features.gypi:31: '<(compiler_dir)/feature_compiler.py',
On 2016/07/14 23:16:36, Dirk Pranke wrote:
> '<(compiler_script)' ?

Done.

[Dirk Pranke, 2016-07-15 19:16:37]

lgtm.

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gni
File build/json_features.gni (right):

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:1: # Copyright 2016 The Chromium Authors. All rights
reserved.
On 2016/07/15 16:19:07, Devlin wrote:
> On 2016/07/14 23:16:36, Dirk Pranke wrote:
> > Where is this template actually going to be used?
> > 
> > We generally don't want to put chromium-specific stuff in //build, as we
want
> to
> > get //build to be a generic things that can be reused by other
(non-chromium)
> > projects.
> > 
> > Can you move these to //tools/json_schema_compiler/json_features.gni or
> > something like that?
> 
> This is going to be used from //chrome and //extensions.
> 
> I just copied where the other similar templates were. :) We can move this, as
> long as there's no qualms about including tools/json_schema_compiler templates
> from chromium sources.
> 
> Done.

including //tools/json_schema_compiler is totally fine.

The location of the templates isn't totally consistent (and we're hopefully
making things better over time), but it is better to keep them by the code where
possible.

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:47: public_deps = lib_public_deps
On 2016/07/15 16:19:07, Devlin wrote:
> On 2016/07/14 23:16:36, Dirk Pranke wrote:
> > similarly, just inline the values here.
> 
> This seems to work for public deps, but trying to inline sources gives an
error:
> ERROR at //build/json_features.gni:43:33: Target not found in this context.
>     sources = invoker.sources + get_target_outputs(":target")
>                                 ^---------------------------

Did you drop the "$" ? I think (but am not positive) that it should work
otherwise.

But not a big deal either way.

[Devlin, 2016-07-15 21:03:01]

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gni
File build/json_features.gni (right):

https://codereview.chromium.org/2151583003/diff/100001/build/json_features.gn...
build/json_features.gni:47: public_deps = lib_public_deps
On 2016/07/15 19:16:37, Dirk Pranke wrote:
> On 2016/07/15 16:19:07, Devlin wrote:
> > On 2016/07/14 23:16:36, Dirk Pranke wrote:
> > > similarly, just inline the values here.
> > 
> > This seems to work for public deps, but trying to inline sources gives an
> error:
> > ERROR at //build/json_features.gni:43:33: Target not found in this context.
> >     sources = invoker.sources + get_target_outputs(":target")
> >                                 ^---------------------------
> 
> Did you drop the "$" ? I think (but am not positive) that it should work
> otherwise.
> 
> But not a big deal either way.

Works now using generated_files var.

[Dirk Pranke, 2016-07-15 21:09:40]

lgtm.

[asargent_no_longer_on_chrome, 2016-07-19 05:33:16]

For features that are only allowed on some OSes, will we be able to avoid
compiling the code for them where they aren't supported?

https://codereview.chromium.org/2151583003/diff/100001/extensions/common/feat...
File extensions/common/features/simple_feature.h (right):

https://codereview.chromium.org/2151583003/diff/100001/extensions/common/feat...
extensions/common/features/simple_feature.h:89: // supported in feature files.
These should only be used this class and in
typo: "only be used this class"

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
File tools/json_schema_compiler/feature_compiler.py (right):

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:74: #     allowed_type_1:
type_definition,
nit: I found the term "type_definition" to be a little confusing here, because
there are several entries where the allowed_type is all that is required and
there is no additional definition required. 

Some ideas for a better term:
"extra_definition"
"extra_restrictions"
"additional_definition"
"extra_properties"

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:83: #   'cpp_map': A map of
strings to C++ enums. When the compiler sees the given
optional: naming this "enum_map" instead of "cpp_map" might add clarity

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:94: #               
Feature::BLESSED_WEB_PAGE_CONTEXT et al for contexts.
I assume the default is allow_all is false unless explicitly set to True?

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:235: (self.name, key, error))
this looks duplicated with the string generated a couple lines above - maybe
worth putting it in a variable?

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:239: """Returns the C++ value for
a given key's python value, or None if the
it's not really clear what a "C++ value" means here (without looking at the
implementation to see what gets returned)- maybe including an example would
help?

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:290: t = self._GetType(v)
nit: I know you can't use a variable named 'type' because that's a reserved
word, but how about 'typ' or 'value_type' or something?

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:316: for sv in v:
nit: sv is a pretty cryptic variable name - is it supposed to mean "single
value"? how about "val" or "single_value" or "item"?

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:340: for key in json.keys():
optional: 'json' is the name of a pretty commonly used package for json parsing
in python, so while it isn't a reserved word, upon seeing "json..." my brain
immediately jumps to json.loads or json.dumps which are the two methods to
parse/dump json strings. You might consider using another variable name like
"parsed_json" or something (since it is an object that was a result of json
parsing)

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:344: self. _ParseKey(key, json,
key_grammar)
nit: is the extra space before the underscore intentional? (or is this just a
rietveld display bug?)

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:414: assert parent, 'No default
parent found for %s' % parent_name
should parent also be a non-list at this point? if so maybe assert that too

https://codereview.chromium.org/2151583003/diff/160001/tools/json_schema_comp...
File tools/json_schema_compiler/feature_compiler.py (right):

https://codereview.chromium.org/2151583003/diff/160001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:528: print(args)
did you mean to leave this in?

[Devlin, 2016-07-19 16:23:36]

On 2016/07/19 05:33:16, Antony Sargent wrote:
> For features that are only allowed on some OSes, will we be able to avoid
> compiling the code for them where they aren't supported?

Neat idea!  Maybe in some cases.  Sometimes features that are specified as one
platform only are still references (or partially references) in cross-platform
code, though, so it will be on a feature-by-feature basis.  I think that's
probably something that can wait for another CL, since it will be
trial-and-error with our feature files.

https://codereview.chromium.org/2151583003/diff/100001/extensions/common/feat...
File extensions/common/features/simple_feature.h (right):

https://codereview.chromium.org/2151583003/diff/100001/extensions/common/feat...
extensions/common/features/simple_feature.h:89: // supported in feature files.
These should only be used this class and in
On 2016/07/19 05:33:16, Antony Sargent wrote:
> typo: "only be used this class"

Done.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
File tools/json_schema_compiler/feature_compiler.py (right):

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:74: #     allowed_type_1:
type_definition,
On 2016/07/19 05:33:16, Antony Sargent wrote:
> nit: I found the term "type_definition" to be a little confusing here, because
> there are several entries where the allowed_type is all that is required and
> there is no additional definition required. 
> 
> Some ideas for a better term:
> "extra_definition"
> "extra_restrictions"
> "additional_definition"
> "extra_properties"
> 

I like properties - how about optional_properties?

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:83: #   'cpp_map': A map of
strings to C++ enums. When the compiler sees the given
On 2016/07/19 05:33:16, Antony Sargent wrote:
> optional: naming this "enum_map" instead of "cpp_map" might add clarity

Done.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:94: #               
Feature::BLESSED_WEB_PAGE_CONTEXT et al for contexts.
On 2016/07/19 05:33:16, Antony Sargent wrote:
> I assume the default is allow_all is false unless explicitly set to True?

Yep, added a note.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:235: (self.name, key, error))
On 2016/07/19 05:33:16, Antony Sargent wrote:
> this looks duplicated with the string generated a couple lines above - maybe
> worth putting it in a variable?
> 

Done.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:239: """Returns the C++ value for
a given key's python value, or None if the
On 2016/07/19 05:33:16, Antony Sargent wrote:
> it's not really clear what a "C++ value" means here (without looking at the
> implementation to see what gets returned)- maybe including an example would
> help?

Done.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:290: t = self._GetType(v)
On 2016/07/19 05:33:16, Antony Sargent wrote:
> nit: I know you can't use a variable named 'type' because that's a reserved
> word, but how about 'typ' or 'value_type' or something?

Done here.  I think in the smaller functions, 't' is probably okay since they're
pretty self-explanatory, but I agree that here it benefits from a bit extra
definition.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:316: for sv in v:
On 2016/07/19 05:33:16, Antony Sargent wrote:
> nit: sv is a pretty cryptic variable name - is it supposed to mean "single
> value"? how about "val" or "single_value" or "item"?

heh sub_value.  Renamed.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:340: for key in json.keys():
On 2016/07/19 05:33:16, Antony Sargent wrote:
> optional: 'json' is the name of a pretty commonly used package for json
parsing
> in python, so while it isn't a reserved word, upon seeing "json..." my brain
> immediately jumps to json.loads or json.dumps which are the two methods to
> parse/dump json strings. You might consider using another variable name like
> "parsed_json" or something (since it is an object that was a result of json
> parsing)
> 

Good idea, done.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:344: self. _ParseKey(key, json,
key_grammar)
On 2016/07/19 05:33:16, Antony Sargent wrote:
> nit: is the extra space before the underscore intentional? (or is this just a
> rietveld display bug?)
> 

Neither intentional nor a display bug, just a typo (and apparently python
doesn't mind).  Fixed.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:414: assert parent, 'No default
parent found for %s' % parent_name
On 2016/07/19 05:33:16, Antony Sargent wrote:
> should parent also be a non-list at this point? if so maybe assert that too
> 
> 

I don't think it's possible to have a list value with a default parent property?
 Unless the grammar was totally messed up with something like:
[{ feature1 }, { feature2 }, 'default_parent']
but then a lot of other stuff would break.

https://codereview.chromium.org/2151583003/diff/160001/tools/json_schema_comp...
File tools/json_schema_compiler/feature_compiler.py (right):

https://codereview.chromium.org/2151583003/diff/160001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:528: print(args)
On 2016/07/19 05:33:16, Antony Sargent wrote:
> did you mean to leave this in?

I meant to leave it in for trybot debugging, but not for landing :)  Removed.

[Devlin, 2016-07-19 16:23:40]

The CQ bit was checked by rdevlin.cronin@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-19 16:24:15]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[asargent_no_longer_on_chrome, 2016-07-19 16:37:18]

lgtm

Re: eliding code on non-supported OSes, I guess you're right that it could be
complicated in some cases, so maybe we'd end up needing some extra properties in
the .json file or something to be able to distinguish the cases where the entire
feature can be elided or not. Anyhow, fine to leave that as a possible
follow-up, and it may or may not actually save all that much binary size.

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
File tools/json_schema_compiler/feature_compiler.py (right):

https://codereview.chromium.org/2151583003/diff/120001/tools/json_schema_comp...
tools/json_schema_compiler/feature_compiler.py:74: #     allowed_type_1:
type_definition,
On 2016/07/19 16:23:36, Devlin wrote:
> On 2016/07/19 05:33:16, Antony Sargent wrote:
> > nit: I found the term "type_definition" to be a little confusing here,
because
> > there are several entries where the allowed_type is all that is required and
> > there is no additional definition required. 
> > 
> > Some ideas for a better term:
> > "extra_definition"
> > "extra_restrictions"
> > "additional_definition"
> > "extra_properties"
> > 
> 
> I like properties - how about optional_properties?

sure, that works

[commit-bot: I haz the power, 2016-07-19 16:42:22]

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

[commit-bot: I haz the power, 2016-07-19 16:42:23]

Dry run: Try jobs failed on following builders:
  android_arm64_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_arm6...)
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[Devlin, 2016-07-19 17:08:12]

The CQ bit was checked by rdevlin.cronin@chromium.org

[Devlin, 2016-07-19 17:08:13]

The patchset sent to the CQ was uploaded after l-g-t-m from dpranke@chromium.org
Link to the patchset: https://codereview.chromium.org/2151583003/#ps180001
(title: "Antony's")

[commit-bot: I haz the power, 2016-07-19 17:08:35]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-19 17:26:20]

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

[commit-bot: I haz the power, 2016-07-19 17:26:22]

Try jobs failed on following builders:
  android_compile_dbg on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_comp...)

[Devlin, 2016-07-19 17:28:18]

The CQ bit was checked by rdevlin.cronin@chromium.org

[commit-bot: I haz the power, 2016-07-19 17:28:45]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-19 17:44:30]

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

[commit-bot: I haz the power, 2016-07-19 17:44:31]

Try jobs failed on following builders:
  android_arm64_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_arm6...)

[Devlin, 2016-07-19 19:06:11]

The CQ bit was checked by rdevlin.cronin@chromium.org

[commit-bot: I haz the power, 2016-07-19 19:07:12]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-19 19:20:28]

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

[commit-bot: I haz the power, 2016-07-19 19:20:29]

Try jobs failed on following builders:
  android_arm64_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_arm6...)
  android_clang_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_clan...)
  android_compile_dbg on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_comp...)

[Devlin, 2016-07-19 22:37:29]

Patchset #8 (id:200001) has been deleted

[Devlin, 2016-07-19 23:02:05]

The CQ bit was checked by rdevlin.cronin@chromium.org

[Devlin, 2016-07-19 23:02:05]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, asargent@chromium.org
Link to the patchset: https://codereview.chromium.org/2151583003/#ps220001
(title: "include fix")

[commit-bot: I haz the power, 2016-07-19 23:02:54]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-19 23:13:57]

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

[commit-bot: I haz the power, 2016-07-19 23:13:59]

Try jobs failed on following builders:
  android_clang_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_clan...)

[Devlin, 2016-07-19 23:21:10]

The CQ bit was checked by rdevlin.cronin@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-19 23:21:52]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Devlin, 2016-07-19 23:38:36]

Patchset #8 (id:220001) has been deleted

[commit-bot: I haz the power, 2016-07-19 23:42:23]

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

[commit-bot: I haz the power, 2016-07-19 23:42:24]

Dry run: Try jobs failed on following builders:
  android_arm64_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_arm6...)
  android_clang_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_clan...)

[Devlin, 2016-07-19 23:53:53]

The CQ bit was checked by rdevlin.cronin@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-19 23:55:15]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-20 02:47:33]

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

[commit-bot: I haz the power, 2016-07-20 02:47:34]

Dry run: This issue passed the CQ dry run.

[Devlin, 2016-07-20 15:09:28]

The CQ bit was checked by rdevlin.cronin@chromium.org

[Devlin, 2016-07-20 15:09:28]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, asargent@chromium.org
Link to the patchset: https://codereview.chromium.org/2151583003/#ps260001
(title: "Include fix")

[commit-bot: I haz the power, 2016-07-20 15:09:59]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-20 16:21:44]

Description was changed from

==========
[Extensions] Add extension feature generation code

Extension features files are json files that contain information about
which APIs or extension features can be used in which contexts.
Currently, these files are parsed at runtime during both browser
startup and renderer process startup, causing a significant delay.

This patch implements the bulk of the functionality to move towards
generating feature files rather than parsing them from the json, which
should hopefully result in a significant speed improvement.

This introduces the feature_compiler.py, which generates the .h and .cc
files for a given set of features.json files, as well as the build rules
to invoke it automatically and tests for it. This does not add generated
files to the production build of chrome yet.

Sample api_feature_provider.h and .cc:
https://gist.github.com/anonymous/9bb094a90810b6050556761c87c204ac

BUG=280286
==========

to

==========
[Extensions] Add extension feature generation code

Extension features files are json files that contain information about
which APIs or extension features can be used in which contexts.
Currently, these files are parsed at runtime during both browser
startup and renderer process startup, causing a significant delay.

This patch implements the bulk of the functionality to move towards
generating feature files rather than parsing them from the json, which
should hopefully result in a significant speed improvement.

This introduces the feature_compiler.py, which generates the .h and .cc
files for a given set of features.json files, as well as the build rules
to invoke it automatically and tests for it. This does not add generated
files to the production build of chrome yet.

Sample api_feature_provider.h and .cc:
https://gist.github.com/anonymous/9bb094a90810b6050556761c87c204ac

BUG=280286
==========

[commit-bot: I haz the power, 2016-07-20 16:21:46]

Committed patchset #9 (id:260001)

[commit-bot: I haz the power, 2016-07-20 16:21:55]

CQ bit was unchecked.

[commit-bot: I haz the power, 2016-07-20 16:25:06]

Description was changed from

==========
[Extensions] Add extension feature generation code

Extension features files are json files that contain information about
which APIs or extension features can be used in which contexts.
Currently, these files are parsed at runtime during both browser
startup and renderer process startup, causing a significant delay.

This patch implements the bulk of the functionality to move towards
generating feature files rather than parsing them from the json, which
should hopefully result in a significant speed improvement.

This introduces the feature_compiler.py, which generates the .h and .cc
files for a given set of features.json files, as well as the build rules
to invoke it automatically and tests for it. This does not add generated
files to the production build of chrome yet.

Sample api_feature_provider.h and .cc:
https://gist.github.com/anonymous/9bb094a90810b6050556761c87c204ac

BUG=280286
==========

to

==========
[Extensions] Add extension feature generation code

Extension features files are json files that contain information about
which APIs or extension features can be used in which contexts.
Currently, these files are parsed at runtime during both browser
startup and renderer process startup, causing a significant delay.

This patch implements the bulk of the functionality to move towards
generating feature files rather than parsing them from the json, which
should hopefully result in a significant speed improvement.

This introduces the feature_compiler.py, which generates the .h and .cc
files for a given set of features.json files, as well as the build rules
to invoke it automatically and tests for it. This does not add generated
files to the production build of chrome yet.

Sample api_feature_provider.h and .cc:
https://gist.github.com/anonymous/9bb094a90810b6050556761c87c204ac

BUG=280286

Committed: https://crrev.com/88155eb86d2b8b8493eb80d09feb639d34dfdcc4
Cr-Commit-Position: refs/heads/master@{#406575}
==========

[commit-bot: I haz the power, 2016-07-20 16:25:08]

Patchset 9 (id:??) landed as
https://crrev.com/88155eb86d2b8b8493eb80d09feb639d34dfdcc4
Cr-Commit-Position: refs/heads/master@{#406575}