This CL adds support for extension in GYP.

This CL adds support for extension in GYP.

This CL adds the extension target type and adds the link flags to
compile .appex extensions.

BUG=gyp:435
R=justincohen@chromium.org, mark@chromium.org, pkl@chromium.org, sdefresne@chromium.org

Committed: https://code.google.com/p/gyp/source/detail?r=1954

[justincohen, 2014-07-07 14:13:58]

https://codereview.chromium.org/376603002/diff/1/pylib/gyp/generator/ninja.py
File pylib/gyp/generator/ninja.py (right):

https://codereview.chromium.org/376603002/diff/1/pylib/gyp/generator/ninja.py...
pylib/gyp/generator/ninja.py:1091: 
nit: blank line

https://codereview.chromium.org/376603002/diff/1/pylib/gyp/xcode_emulation.py
File pylib/gyp/xcode_emulation.py (right):

https://codereview.chromium.org/376603002/diff/1/pylib/gyp/xcode_emulation.py...
pylib/gyp/xcode_emulation.py:807:
'/System/Library/PrivateFrameworks/PlugInKit.framework/PlugInKit')
Maybe a comment on why this has to be hard coded, and maybe tips on how to find
the new path and update this one when inevitably Apple changes the private
framework path somehow

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/ActionExt...
File test/ios/extension/ActionExtension/ActionViewController.h (right):

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/ActionExt...
test/ios/extension/ActionExtension/ActionViewController.h:1: //
wrong header

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/ActionExt...
File test/ios/extension/ActionExtension/ActionViewController.m (right):

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/ActionExt...
test/ios/extension/ActionExtension/ActionViewController.m:2: // 
ActionViewController.m
wrong header on all files.

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/extension...
File test/ios/extension/extension.gyp (right):

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/extension...
test/ios/extension/extension.gyp:76: #	  '-lpkstart
$(SDKROOT)/System/Library/PrivateFrameworks/PlugInKit.framework/PlugInKit
-fapplication-extension -Xlinker -rpath -Xlinker
@executable_path/../../Frameworks '
remove comment

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/extension...
File test/ios/extension/extension.ninja.xcworkspace/contents.xcworkspacedata
(right):

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/extension...
test/ios/extension/extension.ninja.xcworkspace/contents.xcworkspacedata:1: <?xml
version="1.0" encoding="UTF-8"?>
none of this should be checked in.

[olivierrobin, 2014-07-07 14:36:36]

https://codereview.chromium.org/376603002/diff/1/pylib/gyp/generator/ninja.py
File pylib/gyp/generator/ninja.py (right):

https://codereview.chromium.org/376603002/diff/1/pylib/gyp/generator/ninja.py...
pylib/gyp/generator/ninja.py:1091: 
On 2014/07/07 14:13:58, justincohen wrote:
> nit: blank line

Done.

https://codereview.chromium.org/376603002/diff/1/pylib/gyp/xcode_emulation.py
File pylib/gyp/xcode_emulation.py (right):

https://codereview.chromium.org/376603002/diff/1/pylib/gyp/xcode_emulation.py...
pylib/gyp/xcode_emulation.py:807:
'/System/Library/PrivateFrameworks/PlugInKit.framework/PlugInKit')
On 2014/07/07 14:13:58, justincohen wrote:
> Maybe a comment on why this has to be hard coded, and maybe tips on how to
find
> the new path and update this one when inevitably Apple changes the private
> framework path somehow

Done.

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/ActionExt...
File test/ios/extension/ActionExtension/ActionViewController.h (right):

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/ActionExt...
test/ios/extension/ActionExtension/ActionViewController.h:1: //
oups...
done

On 2014/07/07 14:13:58, justincohen wrote:
> wrong header

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/ActionExt...
File test/ios/extension/ActionExtension/ActionViewController.m (right):

https://codereview.chromium.org/376603002/diff/1/test/ios/extension/ActionExt...
test/ios/extension/ActionExtension/ActionViewController.m:2: // 
ActionViewController.m
On 2014/07/07 14:13:58, justincohen wrote:
> wrong header on all files.

Done.

[justincohen, 2014-07-07 16:21:42]

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/Actio...
File test/ios/extension/ActionExtension/Info.plist (right):

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/Info.plist:12:
<string>com.google.chrome.ios.ExtensionContainer.ActionExtension</string>
We shouldn't use this bundle id for gyp tests.  What do we use for the other iOS
tests?

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/Exten...
File test/ios/extension/ExtensionContainer/Info.plist (right):

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/Exten...
test/ios/extension/ExtensionContainer/Info.plist:10:
<string>com.google.chrome.ios.${PRODUCT_NAME:rfc1034identifier}</string>
same here.

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/exten...
File test/ios/extension/extension.ninja.xcworkspace/contents.xcworkspacedata
(right):

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/exten...
test/ios/extension/extension.ninja.xcworkspace/contents.xcworkspacedata:1: <?xml
version="1.0" encoding="UTF-8"?>
This entire folder extension.ninja.xcworkspace should not be included here.

https://codereview.chromium.org/376603002/diff/20001/test/ios/gyptest-extensi...
File test/ios/gyptest-extension.py (right):

https://codereview.chromium.org/376603002/diff/20001/test/ios/gyptest-extensi...
test/ios/gyptest-extension.py:16: test = TestGyp.TestGyp(formats=['ninja'])
Don't we want to test xcode too?

[olivierrobin, 2014-07-08 09:48:18]

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/Actio...
File test/ios/extension/ActionExtension/Info.plist (right):

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/Info.plist:12:
<string>com.google.chrome.ios.ExtensionContainer.ActionExtension</string>
They use com.google.something, so I changed it to com.google.gyptest.extension

On 2014/07/07 16:21:42, justincohen wrote:
> We shouldn't use this bundle id for gyp tests.  What do we use for the other
iOS
> tests?

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/Exten...
File test/ios/extension/ExtensionContainer/Info.plist (right):

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/Exten...
test/ios/extension/ExtensionContainer/Info.plist:10:
<string>com.google.chrome.ios.${PRODUCT_NAME:rfc1034identifier}</string>
On 2014/07/07 16:21:42, justincohen wrote:
> same here.

Done.

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/exten...
File test/ios/extension/extension.ninja.xcworkspace/contents.xcworkspacedata
(right):

https://codereview.chromium.org/376603002/diff/20001/test/ios/extension/exten...
test/ios/extension/extension.ninja.xcworkspace/contents.xcworkspacedata:1: <?xml
version="1.0" encoding="UTF-8"?>
On 2014/07/07 16:21:42, justincohen wrote:
> This entire folder extension.ninja.xcworkspace should not be included here.

Done.

https://codereview.chromium.org/376603002/diff/20001/test/ios/gyptest-extensi...
File test/ios/gyptest-extension.py (right):

https://codereview.chromium.org/376603002/diff/20001/test/ios/gyptest-extensi...
test/ios/gyptest-extension.py:16: test = TestGyp.TestGyp(formats=['ninja'])
On 2014/07/07 16:21:42, justincohen wrote:
> Don't we want to test xcode too?

We cannot test xcode because xcode compilation of extension requires code
signature.

[justincohen, 2014-07-08 21:31:47]

LGTM.

https://codereview.chromium.org/376603002/diff/40001/test/ios/gyptest-extensi...
File test/ios/gyptest-extension.py (right):

https://codereview.chromium.org/376603002/diff/40001/test/ios/gyptest-extensi...
test/ios/gyptest-extension.py:8: Verifies that ios app bundles are built
correctly.
s/bundles/extensions

https://codereview.chromium.org/376603002/diff/40001/test/ios/gyptest-extensi...
test/ios/gyptest-extension.py:22: # Test that the extension is .bundle
s/bundle/appex

[olivierrobin, 2014-07-09 07:35:01]

Thanks

https://codereview.chromium.org/376603002/diff/40001/test/ios/gyptest-extensi...
File test/ios/gyptest-extension.py (right):

https://codereview.chromium.org/376603002/diff/40001/test/ios/gyptest-extensi...
test/ios/gyptest-extension.py:8: Verifies that ios app bundles are built
correctly.
On 2014/07/08 21:31:47, justincohen wrote:
> s/bundles/extensions

Done.

https://codereview.chromium.org/376603002/diff/40001/test/ios/gyptest-extensi...
test/ios/gyptest-extension.py:22: # Test that the extension is .bundle
On 2014/07/08 21:31:47, justincohen wrote:
> s/bundle/appex

Done.

[pkl (ping after 24h if needed), 2014-07-09 08:07:51]

Mostly style nits. I suspect that many of those are from boilerplate code
generated by Xcode, but please fix them anyway. Thanks!

Does this change require iOS SDK 8?

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/generator/ninj...
File pylib/gyp/generator/ninja.py (right):

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/generator/ninj...
pylib/gyp/generator/ninja.py:1053: spec['type'] == 'extension')
how about this:
  spec['type'] in ('executable', 'extension')

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/input.py
File pylib/gyp/input.py (right):

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/input.py#newco...
pylib/gyp/input.py:1715: FULLY_LINKED_TYPES = ('executable', 'loadable_module',
'extension')
This is just a "normal" local variable. It should be just fully_linked_types
(not all caps). I'm aware that there are other cases in this file that has all
caps local variables...they are wrong too (thought you are not expected to fix
them).

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/xcode_emulatio...
File pylib/gyp/xcode_emulation.py (right):

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/xcode_emulatio...
pylib/gyp/xcode_emulation.py:317: self.spec['type'] == 'extension'):
self.spec[type] in (...)

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/xcode_emulatio...
pylib/gyp/xcode_emulation.py:944: if not (self.isIOS and (self.spec['type'] ==
"executable" or
Use single-quote for strings. Also use "in" operator which is a bit more
concise.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
File test/ios/extension/ActionExtension/ActionViewController.m (right):

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/ActionViewController.m:10: 
Just 1 blank line. And is it really necessary to have this empty declaration?

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/ActionViewController.m:15: -
(instancetype)initWithNibName:(NSString *)nibNameOrNil
No space before *. Here and below.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/ActionViewController.m:21: return self;
Can't you just get rid of this boilerplate initializer? And the -viewDidLoad and
-didReceiveMemoryWarning method below?

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/ActionViewController.m:34: // Return any
edited content to the host app.
indent 2.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
File test/ios/extension/ExtensionContainer/AppDelegate.h (right):

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
test/ios/extension/ExtensionContainer/AppDelegate.h:11: 
just 1 blank line.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
File test/ios/extension/ExtensionContainer/AppDelegate.m (right):

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
test/ios/extension/ExtensionContainer/AppDelegate.m:8: 
Just 1 blank line. And fix the style (e.g. "Type*" instead of "Type *") below.
Remove empty method implementation if they are not required by the protocol.

[olivierrobin, 2014-07-09 12:46:28]

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/generator/ninj...
File pylib/gyp/generator/ninja.py (right):

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/generator/ninj...
pylib/gyp/generator/ninja.py:1053: spec['type'] == 'extension')
On 2014/07/09 08:07:50, pklpkl wrote:
> how about this:
>   spec['type'] in ('executable', 'extension')

Done.

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/input.py
File pylib/gyp/input.py (right):

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/input.py#newco...
pylib/gyp/input.py:1715: FULLY_LINKED_TYPES = ('executable', 'loadable_module',
'extension')
On 2014/07/09 08:07:50, pklpkl wrote:
> This is just a "normal" local variable. It should be just fully_linked_types
> (not all caps). I'm aware that there are other cases in this file that has all
> caps local variables...they are wrong too (thought you are not expected to fix
> them).

Done.

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/xcode_emulatio...
File pylib/gyp/xcode_emulation.py (right):

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/xcode_emulatio...
pylib/gyp/xcode_emulation.py:317: self.spec['type'] == 'extension'):
On 2014/07/09 08:07:50, pklpkl wrote:
> self.spec[type] in (...)

Done.

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/xcode_emulatio...
pylib/gyp/xcode_emulation.py:317: self.spec['type'] == 'extension'):
On 2014/07/09 08:07:50, pklpkl wrote:
> self.spec[type] in (...)

Done.

https://codereview.chromium.org/376603002/diff/60001/pylib/gyp/xcode_emulatio...
pylib/gyp/xcode_emulation.py:944: if not (self.isIOS and (self.spec['type'] ==
"executable" or
On 2014/07/09 08:07:50, pklpkl wrote:
> Use single-quote for strings. Also use "in" operator which is a bit more
> concise.

Done.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
File test/ios/extension/ActionExtension/ActionViewController.m (right):

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/ActionViewController.m:10: 
On 2014/07/09 08:07:50, pklpkl wrote:
> Just 1 blank line. And is it really necessary to have this empty declaration?

Done.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/ActionViewController.m:15: -
(instancetype)initWithNibName:(NSString *)nibNameOrNil
On 2014/07/09 08:07:50, pklpkl wrote:
> No space before *. Here and below.

Done.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/ActionViewController.m:21: return self;
On 2014/07/09 08:07:50, pklpkl wrote:
> Can't you just get rid of this boilerplate initializer? And the -viewDidLoad
and
> -didReceiveMemoryWarning method below?

Done.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Actio...
test/ios/extension/ActionExtension/ActionViewController.m:34: // Return any
edited content to the host app.
On 2014/07/09 08:07:50, pklpkl wrote:
> indent 2.

Done.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
File test/ios/extension/ExtensionContainer/AppDelegate.h (right):

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
test/ios/extension/ExtensionContainer/AppDelegate.h:11: 
On 2014/07/09 08:07:50, pklpkl wrote:
> just 1 blank line.

Done.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
File test/ios/extension/ExtensionContainer/AppDelegate.m (right):

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
test/ios/extension/ExtensionContainer/AppDelegate.m:8: 
On 2014/07/09 08:07:50, pklpkl wrote:
> Just 1 blank line. And fix the style (e.g. "Type*" instead of "Type *") below.
> Remove empty method implementation if they are not required by the protocol.

Done.

https://codereview.chromium.org/376603002/diff/60001/test/ios/extension/Exten...
test/ios/extension/ExtensionContainer/AppDelegate.m:8: 
On 2014/07/09 08:07:50, pklpkl wrote:
> Just 1 blank line. And fix the style (e.g. "Type*" instead of "Type *") below.
> Remove empty method implementation if they are not required by the protocol.

Done.

https://codereview.chromium.org/376603002/diff/60001/test/ios/gyptest-extensi...
File test/ios/gyptest-extension.py (right):

https://codereview.chromium.org/376603002/diff/60001/test/ios/gyptest-extensi...
test/ios/gyptest-extension.py:3: # Copyright (c) 2012 Google Inc. All rights
reserved.
copyright

[pkl (ping after 24h if needed), 2014-07-09 12:53:13]

lgtm

[olivierrobin, 2014-07-09 13:57:46]

From mmentovai :
OK, then I think we should just treat this as 'type': 'executable', and you
should signal that it’s an appex with a platform-specific attribute like
'ios_app_extension': 1 along with 'mac_bundle': 1. Note that this is exactly how
we handle the 'mac_bundle' attribute: it’s not conceptually a different target
type from the rest of GYP’s perspective, it’s a platform-specific attribute that
might have a slight impact on packaging or linking. The semantics for
'executable' should already be correct for appexes with respect to how they’re
linkables that can’t be linked against and how they behave within the dependency
tree.

[olivierrobin, 2014-07-09 15:27:12]

I changed the flag name to ios_app_extension, accordingly to mmentovai advice

[Mark Mentovai, 2014-07-09 18:02:52]

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/generator/xco...
File pylib/gyp/generator/xcode.py (right):

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/generator/xco...
pylib/gyp/generator/xcode.py:676: type_bundle_key += '+extension+bundle'
You should require (assert is OK) that is_bundle is true in here.

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
File pylib/gyp/xcode_emulation.py (right):

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
pylib/gyp/xcode_emulation.py:222: return int(self.spec.get('ios_app_extension',
0)) != 0
You can require (assert) that _IsBundle is true if this returns true, also. Or
you may wish to do check it in GetProductType at line 302.

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
pylib/gyp/xcode_emulation.py:810: ldflags.append('-lpkstart ' + sdk_root +
Some of the other flags in here set a bad example, but let’s not continue the
practice.

-lpkstart should be a separate argument from os.path.join(sdk_root,
'…path/to/PlugInKit').

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
pylib/gyp/xcode_emulation.py:812: ldflags.append('-fapplication-extension ' +
And these should be 5 separate arguments.

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcodeproj_fil...
File pylib/gyp/xcodeproj_file.py (right):

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcodeproj_fil...
pylib/gyp/xcodeproj_file.py:2244: 'com.apple.product-type.app-extension':    
['wrapper.application',
Is wrapper.application really what Xcode uses?

https://codereview.chromium.org/376603002/diff/100001/test/ios/extension/Acti...
File test/ios/extension/ActionExtension/ActionViewController.h (right):

https://codereview.chromium.org/376603002/diff/100001/test/ios/extension/Acti...
test/ios/extension/ActionExtension/ActionViewController.h:1: // Copyright (c)
2014 Google Inc. All rights reserved.
Can you adjust your --similarity so that these don’t look like file copies?
They’re not file copies.

https://codereview.chromium.org/376603002/diff/100001/test/ios/extension/Exte...
File test/ios/extension/ExtensionContainer/main.m (right):

https://codereview.chromium.org/376603002/diff/100001/test/ios/extension/Exte...
test/ios/extension/ExtensionContainer/main.m:8: int main(int argc, char *
argv[]) {
Snuggle the * up against the char.

https://codereview.chromium.org/376603002/diff/100001/test/ios/extension/Exte...
test/ios/extension/ExtensionContainer/main.m:10: return UIApplicationMain(argc,
argv, nil,
Weird indentation.

[olivierrobin, 2014-07-10 09:31:51]

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/generator/xco...
File pylib/gyp/generator/xcode.py (right):

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/generator/xco...
pylib/gyp/generator/xcode.py:676: type_bundle_key += '+extension+bundle'
On 2014/07/09 18:02:51, Mark Mentovai wrote:
> You should require (assert is OK) that is_bundle is true in here.

Done.

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
File pylib/gyp/xcode_emulation.py (right):

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
pylib/gyp/xcode_emulation.py:222: return int(self.spec.get('ios_app_extension',
0)) != 0
Done on line 302

On 2014/07/09 18:02:52, Mark Mentovai wrote:
> You can require (assert) that _IsBundle is true if this returns true, also. Or
> you may wish to do check it in GetProductType at line 302.

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
pylib/gyp/xcode_emulation.py:810: ldflags.append('-lpkstart ' + sdk_root +
On 2014/07/09 18:02:52, Mark Mentovai wrote:
> Some of the other flags in here set a bad example, but let’s not continue the
> practice.
> 
> -lpkstart should be a separate argument from os.path.join(sdk_root,
> '…path/to/PlugInKit').

Done.

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
pylib/gyp/xcode_emulation.py:810: ldflags.append('-lpkstart ' + sdk_root +
On 2014/07/09 18:02:52, Mark Mentovai wrote:
> Some of the other flags in here set a bad example, but let’s not continue the
> practice.
> 
> -lpkstart should be a separate argument from os.path.join(sdk_root,
> '…path/to/PlugInKit').

Done.

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcode_emulati...
pylib/gyp/xcode_emulation.py:812: ldflags.append('-fapplication-extension ' +
On 2014/07/09 18:02:51, Mark Mentovai wrote:
> And these should be 5 separate arguments.

I separated fapplication-extension from the others, but it seems that you cannot
have twice the same flag, meaning that if I separate both -Xlinker, I get it
only once and the linking fail.

As you may have more than one rpath, it seems required to put the four arguments
in the same flag.

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcodeproj_fil...
File pylib/gyp/xcodeproj_file.py (right):

https://codereview.chromium.org/376603002/diff/100001/pylib/gyp/xcodeproj_fil...
pylib/gyp/xcodeproj_file.py:2244: 'com.apple.product-type.app-extension':    
['wrapper.application',
You are right, sorry. done
On 2014/07/09 18:02:52, Mark Mentovai wrote:
> Is wrapper.application really what Xcode uses?

https://codereview.chromium.org/376603002/diff/100001/test/ios/extension/Exte...
File test/ios/extension/ExtensionContainer/main.m (right):

https://codereview.chromium.org/376603002/diff/100001/test/ios/extension/Exte...
test/ios/extension/ExtensionContainer/main.m:8: int main(int argc, char *
argv[]) {
On 2014/07/09 18:02:52, Mark Mentovai wrote:
> Snuggle the * up against the char.

Done.

https://codereview.chromium.org/376603002/diff/100001/test/ios/extension/Exte...
test/ios/extension/ExtensionContainer/main.m:10: return UIApplicationMain(argc,
argv, nil,
On 2014/07/09 18:02:52, Mark Mentovai wrote:
> Weird indentation.

Done.

[olivierrobin, 2014-07-15 14:09:41]

Hi mark

Can you take another look and tell me if my modifications are ok ?

Thank you

Olivier Robin

[Mark Mentovai, 2014-07-15 21:49:35]

LGTM

[olivierrobin, 2014-07-23 08:28:16]

+sdefresne for upstream commit

[sdefresne, 2014-07-24 08:06:16]

Updated the description to point to the bug.

[sdefresne, 2014-07-24 08:06:20]

lgtm

Updated the description to point to the bug.

[sdefresne, 2014-07-24 08:11:13]

Committed patchset #8 manually as r1954 (presubmit successful).