Add support for references in different paths in apis

- Add support for references in different paths in apis.

- Move ImageDetails from chrome to extensions.

BUG=352290
Committed: https://crrev.com/8a1bee36d9f609ab9a98ecf89bf1dfc0bb4762dc
Cr-Commit-Position: refs/heads/master@{#292567}

[lfg, 2014-08-25 21:02:22]

Hi kalman,

This is a first pass on adding support for references in different paths, please
take a look on all files.

[not at google - send to devlin, 2014-08-25 22:22:47]

Awesome work untangling that.

Making that "environment" change will improve dependency injection and mean you
won't have to change so many files.

That that means for schema_loader.py -- I don't know -- hard to see until the
change is made.

https://codereview.chromium.org/487533005/diff/20001/build/json_schema_bundle...
File build/json_schema_bundle_compile.gypi (right):

https://codereview.chromium.org/487533005/diff/20001/build/json_schema_bundle...
build/json_schema_bundle_compile.gypi:25: 'schema_includes' : [],
No space before the :

https://codereview.chromium.org/487533005/diff/20001/build/json_schema_bundle...
build/json_schema_bundle_compile.gypi:60: '--include=<(schema_includes)',
For every time this parameter is passed around, or some form of it:

Can we standardise on a single terminology, whether that be "include" or
"schema-include" or, whatever? I think "include" by itself is too vague,
"include-dir" more accurate but doesn't capture the way that these aren't
actually directories, they're rules which include the namespace with which to
include.

I wish there was a way for gyp to do all of this for us, but, ah well.

So how about "include-rules"?

https://codereview.chromium.org/487533005/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/api.gyp (right):

https://codereview.chromium.org/487533005/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/api.gyp:124: 'main_schema_includes': [
-include-dirs?

https://codereview.chromium.org/487533005/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/tabs.json (right):

https://codereview.chromium.org/487533005/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/tabs.json:657: "$ref":
"extension_types.ImageDetails",
ideally this should be extesionTypes.ImageDetails. API namespaces are always
camelcase.

https://codereview.chromium.org/487533005/diff/20001/extensions/common/api/ex...
File extensions/common/api/extension_types.json (right):

https://codereview.chromium.org/487533005/diff/20001/extensions/common/api/ex...
extensions/common/api/extension_types.json:7: "namespace": "extension_types",
extensionTypes

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
File tools/json_schema_compiler/compiler.py (right):

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:22: import shlex
Alphabetic order.

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:155: help='A list of paths to include
when searching for referenced objects')
Comment should also mention that convention of : separation

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:172: path, namespace = i.split(':', 1)
This assumes that there is a ':' in the name. Maybe there should be, in which
case we should do some validation so that the error people see when they forget
to add a ':' isn't "ValueError: need more than 1 value to unpack". Suggestion
below:

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:173: includes.append( { 'path': path,
'namespace': namespace } )
How about:

def split_path_and_namespace(path_and_namespace):
  if ':' not in path_and_namespace:
    raise ValueError('Invalid include rule "%s". Rules must be of '
                     'the form path:namespace' % path_and_namespace)
  return path_and_namespace.split(':', 1)

includes [{'path': path, 'namespace': namespace},
          for path, namespace in split_path_and_namespace(rule)
          for rule in shlex.split(opts.include)]

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
File tools/json_schema_compiler/model.py (right):

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/model.py:33: cpp_namespace_pattern,
There's a reason a lot of these namespace-level properties aren't passed into
Namespace - it's supposed to be language agnostic. Pure model. Compiler specific
options should go in "compiler_options" but that's supposed to be configured at
the schema level, a literal replacement of the "compiler_options" schema
argument.

How about:

- add an optional "environment" variable to this which you can start bundling
those namespace-level options into. Ideally it'd even be typed, like,

class CppNamespaceEnvironment(object):
  def __init__(self, namespace_pattern):
    self.namespace_pattern = namespace_pattern

...

class Model(object):
  def AddNamespace(self, json, source_file,
                   include_compiler_options=False,
                   environment=None)
  ...

...

model.AddNamespace(
    json,
    source_file,
    environment=CppNamespaceEnvironment('foo::bar::baz'))

...

then access like "namespace.environment.namespace_pattern".

[lfg, 2014-08-26 16:15:22]

https://codereview.chromium.org/487533005/diff/20001/build/json_schema_bundle...
File build/json_schema_bundle_compile.gypi (right):

https://codereview.chromium.org/487533005/diff/20001/build/json_schema_bundle...
build/json_schema_bundle_compile.gypi:25: 'schema_includes' : [],
On 2014/08/25 22:22:46, kalman wrote:
> No space before the :

Done.

https://codereview.chromium.org/487533005/diff/20001/build/json_schema_bundle...
build/json_schema_bundle_compile.gypi:60: '--include=<(schema_includes)',
On 2014/08/25 22:22:46, kalman wrote:
> For every time this parameter is passed around, or some form of it:
> 
> Can we standardise on a single terminology, whether that be "include" or
> "schema-include" or, whatever? I think "include" by itself is too vague,
> "include-dir" more accurate but doesn't capture the way that these aren't
> actually directories, they're rules which include the namespace with which to
> include.
> 
> I wish there was a way for gyp to do all of this for us, but, ah well.
> 
> So how about "include-rules"?

include-rules sounds good, I went with it.

https://codereview.chromium.org/487533005/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/api.gyp (right):

https://codereview.chromium.org/487533005/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/api.gyp:124: 'main_schema_includes': [
On 2014/08/25 22:22:46, kalman wrote:
> -include-dirs?

Changed to include_rules, to use the same standard as in the compiler.

https://codereview.chromium.org/487533005/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/tabs.json (right):

https://codereview.chromium.org/487533005/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/tabs.json:657: "$ref":
"extension_types.ImageDetails",
On 2014/08/25 22:22:47, kalman wrote:
> ideally this should be extesionTypes.ImageDetails. API namespaces are always
> camelcase.

Done.

https://codereview.chromium.org/487533005/diff/20001/extensions/common/api/ex...
File extensions/common/api/extension_types.json (right):

https://codereview.chromium.org/487533005/diff/20001/extensions/common/api/ex...
extensions/common/api/extension_types.json:7: "namespace": "extension_types",
On 2014/08/25 22:22:47, kalman wrote:
> extensionTypes

Done.

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
File tools/json_schema_compiler/compiler.py (right):

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:22: import shlex
On 2014/08/25 22:22:47, kalman wrote:
> Alphabetic order.

Done.

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:155: help='A list of paths to include
when searching for referenced objects')
On 2014/08/25 22:22:47, kalman wrote:
> Comment should also mention that convention of : separation

Done.

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:172: path, namespace = i.split(':', 1)
On 2014/08/25 22:22:47, kalman wrote:
> This assumes that there is a ':' in the name. Maybe there should be, in which
> case we should do some validation so that the error people see when they
forget
> to add a ':' isn't "ValueError: need more than 1 value to unpack". Suggestion
> below:

Done.

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:173: includes.append( { 'path': path,
'namespace': namespace } )
On 2014/08/25 22:22:47, kalman wrote:
> How about:
> 
> def split_path_and_namespace(path_and_namespace):
>   if ':' not in path_and_namespace:
>     raise ValueError('Invalid include rule "%s". Rules must be of '
>                      'the form path:namespace' % path_and_namespace)
>   return path_and_namespace.split(':', 1)
> 
> includes [{'path': path, 'namespace': namespace},
>           for path, namespace in split_path_and_namespace(rule)
>           for rule in shlex.split(opts.include)]

Done.

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
File tools/json_schema_compiler/model.py (right):

https://codereview.chromium.org/487533005/diff/20001/tools/json_schema_compil...
tools/json_schema_compiler/model.py:33: cpp_namespace_pattern,
On 2014/08/25 22:22:47, kalman wrote:
> There's a reason a lot of these namespace-level properties aren't passed into
> Namespace - it's supposed to be language agnostic. Pure model. Compiler
specific
> options should go in "compiler_options" but that's supposed to be configured
at
> the schema level, a literal replacement of the "compiler_options" schema
> argument.
> 
> How about:
> 
> - add an optional "environment" variable to this which you can start bundling
> those namespace-level options into. Ideally it'd even be typed, like,
> 
> class CppNamespaceEnvironment(object):
>   def __init__(self, namespace_pattern):
>     self.namespace_pattern = namespace_pattern
> 
> ...
> 
> class Model(object):
>   def AddNamespace(self, json, source_file,
>                    include_compiler_options=False,
>                    environment=None)
>   ...
> 
> ...
> 
> model.AddNamespace(
>     json,
>     source_file,
>     environment=CppNamespaceEnvironment('foo::bar::baz'))
> 
> ...
> 
> then access like "namespace.environment.namespace_pattern".

I thought this didn't feel right when I had to fix some of the tests and the
cpp_namespace_pattern didn't make sense.

Fixed with your suggestion.

[not at google - send to devlin, 2014-08-26 20:16:06]

Next round of comments. Looking much better, thanks.

Unfortunately this breaks the documentation for captureVisibleTab because
ImageDetails can no longer be found. I'm still looking into this, but it's
related to crbug.com/402763.

https://codereview.chromium.org/487533005/diff/40001/build/json_schema_bundle...
File build/json_schema_bundle_compile.gypi (right):

https://codereview.chromium.org/487533005/diff/40001/build/json_schema_bundle...
build/json_schema_bundle_compile.gypi:10: #   schema_includes (optional):
Please call this "schema_include_rules".

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/cc_generator.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/cc_generator.py:33: if self._namespace.environment:
When will there not be an environment? Could you assert that there's an
environment in CCGenerator's constructor?

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/compiler.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:156: parser.add_option('-I',
'--include-rules', #dest='include',
Remove #dest comment

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:180: for path, namespace in [
split_path_and_namespace(rule) ]]
Ah ok. So writing this singleton list thing is a bit odd. Maybe that
split_path_and_namespace function should be returning that {'path': path,
'namespace': namespace} object rather than a tuple. Then you could even write
like:

def parse_path_and_namspace(path_and_namespace):
  ...
  path, namespace = path_and_namespace.split(':', 1)
  return {'path': path, 'namespace': namespace}
include_rules = map(parse_path_and_namspace,
                    shlex.split(opts.include_rules))

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:184: opts.impl_dir, includes)
(call this include_rules)

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/cpp_type_generator.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/cpp_type_generator.py:25: if namespace.environment:
Again, when will there not be an environment? I would have thought that all cpp
generators would have a cpp environment, and this file should assume it's only
being called for cpp generators.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/cpp_type_generator_test.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/cpp_type_generator_test.py:106:
self.browser_action_json[0], 'path/to/browser_action.json')
Put this back as 1 line?

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/cpp_type_generator_test.py:116: '',
manager.GenerateForwardDeclarations().Render())
Can this be a single line?

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/h_generator.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/h_generator.py:66: if self._namespace.environment:
(Same comment re lack of environment)

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/schema_loader.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/schema_loader.py:58: for include in self._includes:
You should be able to write here:

for path, namespace in self._includes.iteritems():
  ...

and not need to pull out properties of |include| like on lines 59 and 66.

[lfg, 2014-08-26 21:18:44]

Here's a new pass with some fixes.

I also don't know why linux_chromium_gn_rel trybot is failing (it's not a
flaky), if you have any idea, let me know.

https://codereview.chromium.org/487533005/diff/40001/build/json_schema_bundle...
File build/json_schema_bundle_compile.gypi (right):

https://codereview.chromium.org/487533005/diff/40001/build/json_schema_bundle...
build/json_schema_bundle_compile.gypi:10: #   schema_includes (optional):
On 2014/08/26 20:16:05, kalman wrote:
> Please call this "schema_include_rules".

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/cc_generator.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/cc_generator.py:33: if self._namespace.environment:
On 2014/08/26 20:16:06, kalman wrote:
> When will there not be an environment? Could you assert that there's an
> environment in CCGenerator's constructor?

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/compiler.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:156: parser.add_option('-I',
'--include-rules', #dest='include',
On 2014/08/26 20:16:06, kalman wrote:
> Remove #dest comment

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:180: for path, namespace in [
split_path_and_namespace(rule) ]]
On 2014/08/26 20:16:06, kalman wrote:
> Ah ok. So writing this singleton list thing is a bit odd. Maybe that
> split_path_and_namespace function should be returning that {'path': path,
> 'namespace': namespace} object rather than a tuple. Then you could even write
> like:
> 
> def parse_path_and_namspace(path_and_namespace):
>   ...
>   path, namespace = path_and_namespace.split(':', 1)
>   return {'path': path, 'namespace': namespace}
> include_rules = map(parse_path_and_namspace,
>                     shlex.split(opts.include_rules))

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:184: opts.impl_dir, includes)
On 2014/08/26 20:16:06, kalman wrote:
> (call this include_rules)

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/cpp_type_generator.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/cpp_type_generator.py:25: if namespace.environment:
On 2014/08/26 20:16:06, kalman wrote:
> Again, when will there not be an environment? I would have thought that all
cpp
> generators would have a cpp environment, and this file should assume it's only
> being called for cpp generators.

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/cpp_type_generator_test.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/cpp_type_generator_test.py:106:
self.browser_action_json[0], 'path/to/browser_action.json')
On 2014/08/26 20:16:06, kalman wrote:
> Put this back as 1 line?

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/cpp_type_generator_test.py:116: '',
manager.GenerateForwardDeclarations().Render())
On 2014/08/26 20:16:06, kalman wrote:
> Can this be a single line?

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/h_generator.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/h_generator.py:66: if self._namespace.environment:
On 2014/08/26 20:16:06, kalman wrote:
> (Same comment re lack of environment)

Done.

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/schema_loader.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/schema_loader.py:58: for include in self._includes:
On 2014/08/26 20:16:06, kalman wrote:
> You should be able to write here:
> 
> for path, namespace in self._includes.iteritems():
>   ...
> 
> and not need to pull out properties of |include| like on lines 59 and 66.

I don't think this can be done with dictionaries, as it'll get key/value pairs.
I've switched to tuples to be able to do this.

[not at google - send to devlin, 2014-08-27 23:11:23]

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/schema_loader.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/schema_loader.py:58: for include in self._includes:
On 2014/08/26 21:18:44, lfg wrote:
> On 2014/08/26 20:16:06, kalman wrote:
> > You should be able to write here:
> > 
> > for path, namespace in self._includes.iteritems():
> >   ...
> > 
> > and not need to pull out properties of |include| like on lines 59 and 66.
> 
> I don't think this can be done with dictionaries, as it'll get key/value
pairs.
> I've switched to tuples to be able to do this.

That's what iteritems() does, turns a dictionary into key/value pairs.

[not at google - send to devlin, 2014-08-27 23:14:21]

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/schema_loader.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/schema_loader.py:58: for include in self._includes:
On 2014/08/27 23:11:23, kalman wrote:
> On 2014/08/26 21:18:44, lfg wrote:
> > On 2014/08/26 20:16:06, kalman wrote:
> > > You should be able to write here:
> > > 
> > > for path, namespace in self._includes.iteritems():
> > >   ...
> > > 
> > > and not need to pull out properties of |include| like on lines 59 and 66.
> > 
> > I don't think this can be done with dictionaries, as it'll get key/value
> pairs.
> > I've switched to tuples to be able to do this.
> 
> That's what iteritems() does, turns a dictionary into key/value pairs.

(but tuples is fine as well)

https://codereview.chromium.org/487533005/diff/60001/tools/json_schema_compil...
File tools/json_schema_compiler/cc_generator.py (right):

https://codereview.chromium.org/487533005/diff/60001/tools/json_schema_compil...
tools/json_schema_compiler/cc_generator.py:24: raise "CppNamespaceEnvironment
not defined"
When I said assert I literally meant call "assert".

https://codereview.chromium.org/487533005/diff/60001/tools/json_schema_compil...
File tools/json_schema_compiler/compiler.py (right):

https://codereview.chromium.org/487533005/diff/60001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:177: return (path, namespace)
You can just return path_and_namespace.split(':', 1) - no need to assign it to
these temporary variables.

[not at google - send to devlin, 2014-08-27 23:16:55]

kalman@chromium.org changed reviewers:
+ rockot@chromium.org

[not at google - send to devlin, 2014-08-27 23:16:55]

Looks good - but is there any chance you're able to look at the docs problem
soon? Or Ken, you triage it? This is going to make the docs situation worse, and
I don't have the bandwidth to fix it myself.

[lfg, 2014-08-28 18:57:10]

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
File tools/json_schema_compiler/schema_loader.py (right):

https://codereview.chromium.org/487533005/diff/40001/tools/json_schema_compil...
tools/json_schema_compiler/schema_loader.py:58: for include in self._includes:
On 2014/08/27 23:14:20, kalman wrote:
> On 2014/08/27 23:11:23, kalman wrote:
> > On 2014/08/26 21:18:44, lfg wrote:
> > > On 2014/08/26 20:16:06, kalman wrote:
> > > > You should be able to write here:
> > > > 
> > > > for path, namespace in self._includes.iteritems():
> > > >   ...
> > > > 
> > > > and not need to pull out properties of |include| like on lines 59 and
66.
> > > 
> > > I don't think this can be done with dictionaries, as it'll get key/value
> > pairs.
> > > I've switched to tuples to be able to do this.
> > 
> > That's what iteritems() does, turns a dictionary into key/value pairs.
> 
> (but tuples is fine as well)

Done.

https://codereview.chromium.org/487533005/diff/60001/tools/json_schema_compil...
File tools/json_schema_compiler/cc_generator.py (right):

https://codereview.chromium.org/487533005/diff/60001/tools/json_schema_compil...
tools/json_schema_compiler/cc_generator.py:24: raise "CppNamespaceEnvironment
not defined"
On 2014/08/27 23:14:20, kalman wrote:
> When I said assert I literally meant call "assert".

Done.

https://codereview.chromium.org/487533005/diff/60001/tools/json_schema_compil...
File tools/json_schema_compiler/compiler.py (right):

https://codereview.chromium.org/487533005/diff/60001/tools/json_schema_compil...
tools/json_schema_compiler/compiler.py:177: return (path, namespace)
On 2014/08/27 23:14:20, kalman wrote:
> You can just return path_and_namespace.split(':', 1) - no need to assign it to
> these temporary variables.

Done.

[not at google - send to devlin, 2014-08-28 19:42:39]

lgtm

[lfg, 2014-08-28 19:45:27]

The CQ bit was checked by lfg@chromium.org

[commit-bot: I haz the power, 2014-08-28 19:49:07]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/lfg@chromium.org/487533005/140001

[lfg, 2014-08-28 20:12:11]

The CQ bit was unchecked by lfg@chromium.org

[lfg, 2014-08-28 22:57:31]

lfg@chromium.org changed reviewers:
+ yoz@chromium.org

[lfg, 2014-08-28 22:57:31]

@kalman:

PTAL, I had to change the way paths are accessed to fix the gn build

@yoz:

Please, look at the .gn/.gni files.

[not at google - send to devlin, 2014-08-28 23:14:28]

lgtm

[Yoyo Zhou, 2014-08-28 23:17:21]

Mostly OK but

https://codereview.chromium.org/487533005/diff/180001/chrome/common/extension...
File chrome/common/extensions/api/api.gyp (right):

https://codereview.chromium.org/487533005/diff/180001/chrome/common/extension...
chrome/common/extensions/api/api.gyp:21: 'dependencies': [
I expect to see this reflected in the BUILD.gn.

[lfg, 2014-08-29 00:45:13]

https://codereview.chromium.org/487533005/diff/180001/chrome/common/extension...
File chrome/common/extensions/api/api.gyp (right):

https://codereview.chromium.org/487533005/diff/180001/chrome/common/extension...
chrome/common/extensions/api/api.gyp:21: 'dependencies': [
On 2014/08/28 23:17:21, Yoyo Zhou wrote:
> I expect to see this reflected in the BUILD.gn.

Done.

[Yoyo Zhou, 2014-08-29 00:56:14]

LGTM

https://codereview.chromium.org/487533005/diff/200001/chrome/common/extension...
File chrome/common/extensions/api/schemas.gni (right):

https://codereview.chromium.org/487533005/diff/200001/chrome/common/extension...
chrome/common/extensions/api/schemas.gni:34: schema_dependencies = [
"//extensions/common/api:api" ]
nit: you don't need the :api here

[lfg, 2014-08-29 02:00:30]

The CQ bit was checked by lfg@chromium.org

[commit-bot: I haz the power, 2014-08-29 02:01:41]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/lfg@chromium.org/487533005/220001

[commit-bot: I haz the power, 2014-08-29 03:06:52]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  win_chromium_rel_swarming on tryserver.chromium.win
(http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_rel_...)

[commit-bot: I haz the power, 2014-08-29 03:56:36]

Committed patchset #12 (id:220001) as 2647554914925ef62da6b359d22e5e7ce584dd89

[commit-bot: I haz the power, 2014-09-10 03:06:07]

Patchset 12 (id:??) landed as
https://crrev.com/8a1bee36d9f609ab9a98ecf89bf1dfc0bb4762dc
Cr-Commit-Position: refs/heads/master@{#292567}