Rewrite most of the scopers in //base/mac to use ScopedTypeRef or ScopedGeneric.

Rewrite most of the scopers in //base/mac to use ScopedTypeRef or ScopedGeneric.

This removes a lot of duplicated code in favor of using type aliases and
minimal Traits structs.

R=mark@chromium.org

Committed: https://chromium.googlesource.com/chromium/src/+/cfd6ed5600d810b695f651a1d6f52ea62d397d1c

[Robert Sesek, 2015-12-30 21:36:07]

Patchset #2 (id:20001) has been deleted

[Mark Mentovai, 2016-01-04 16:09:24]

LGTM

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_typeref.h
File base/mac/scoped_typeref.h (right):

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_typeref...
base/mac/scoped_typeref.h:25: //     static CGLContextObj Retain(CGLContextObj
object) {
I think it’s easier to have these return void and consolidate the object_ =
object setting all in this one file, rather than making all of the traits have
to use an extra line. In practice, the extra line in the traits turns into three
extra lines as it did in this example.

Simpler traits are better than more-elegant internals, and I don’t think
you have any traits where the retain function returns something other than its
parameter anyway.

[Robert Sesek, 2016-01-04 16:11:53]

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_typeref.h
File base/mac/scoped_typeref.h (right):

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_typeref...
base/mac/scoped_typeref.h:25: //     static CGLContextObj Retain(CGLContextObj
object) {
On 2016/01/04 16:09:23, Mark Mentovai wrote:
> I think it’s easier to have these return void and consolidate the object_ =
> object setting all in this one file, rather than making all of the traits have
> to use an extra line. In practice, the extra line in the traits turns into
three
> extra lines as it did in this example.
> 
> Simpler traits are better than more-elegant internals, and I don’t think
> you have any traits where the retain function returns something other
than its
> parameter anyway.

I had to make this change for ScopedBlock, since there's only Block_copy (which
creates a new heap block, potentially from a stack block) not a Block_retain.

[Mark Mentovai, 2016-01-04 16:11:56]

ScopedBlock. OK. LGTM.

[commit-bot: I haz the power, 2016-01-04 17:45:10]

Description was changed from

==========
Rewrite most of the scopers in //base/mac to use ScopedTypeRef or ScopedGeneric.

This removes a lot of duplicated code in favor of using type aliases and
minimal Traits structs.

R=mark@chromium.org
==========

to

==========
Rewrite most of the scopers in //base/mac to use ScopedTypeRef or ScopedGeneric.

This removes a lot of duplicated code in favor of using type aliases and
minimal Traits structs.

R=mark@chromium.org

Committed: https://crrev.com/cfd6ed5600d810b695f651a1d6f52ea62d397d1c
Cr-Commit-Position: refs/heads/master@{#367322}
==========

[commit-bot: I haz the power, 2016-01-04 17:45:11]

Patchset 2 (id:??) landed as
https://crrev.com/cfd6ed5600d810b695f651a1d6f52ea62d397d1c
Cr-Commit-Position: refs/heads/master@{#367322}

[Robert Sesek, 2016-01-04 17:45:48]

Description was changed from

==========
Rewrite most of the scopers in //base/mac to use ScopedTypeRef or ScopedGeneric.

This removes a lot of duplicated code in favor of using type aliases and
minimal Traits structs.

R=mark@chromium.org

Committed: https://crrev.com/cfd6ed5600d810b695f651a1d6f52ea62d397d1c
Cr-Commit-Position: refs/heads/master@{#367322}
==========

to

==========
Rewrite most of the scopers in //base/mac to use ScopedTypeRef or ScopedGeneric.

This removes a lot of duplicated code in favor of using type aliases and
minimal Traits structs.

R=mark@chromium.org

Committed:
https://chromium.googlesource.com/chromium/src/+/cfd6ed5600d810b695f651a1d6f5...
==========

[Robert Sesek, 2016-01-04 17:45:49]

Committed patchset #2 (id:40001) manually as
cfd6ed5600d810b695f651a1d6f52ea62d397d1c.

[Nico, 2016-01-05 16:18:29]

thakis@chromium.org changed reviewers:
+ thakis@chromium.org

[Nico, 2016-01-05 16:18:30]

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_nsobjec...
File base/mac/scoped_nsobject_unittest.mm (right):

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_nsobjec...
base/mac/scoped_nsobject_unittest.mm:55: #if 0
?

[Robert Sesek, 2016-01-05 16:26:10]

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_nsobjec...
File base/mac/scoped_nsobject_unittest.mm (right):

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_nsobjec...
base/mac/scoped_nsobject_unittest.mm:55: #if 0
On 2016/01/05 16:18:30, Nico wrote:
> ?

It's for testing a static_assert.

[Nico, 2016-01-05 16:38:40]

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_nsobjec...
File base/mac/scoped_nsobject_unittest.mm (right):

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_nsobjec...
base/mac/scoped_nsobject_unittest.mm:55: #if 0
On 2016/01/05 16:26:10, Robert Sesek wrote:
> On 2016/01/05 16:18:30, Nico wrote:
> > ?
> 
> It's for testing a static_assert.

This should have a comment saying so, else it's be deleted by the next person
touching this file. Or it should be in a .nc file.

[Robert Sesek, 2016-01-05 17:35:03]

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_nsobjec...
File base/mac/scoped_nsobject_unittest.mm (right):

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_nsobjec...
base/mac/scoped_nsobject_unittest.mm:55: #if 0
On 2016/01/05 16:38:39, Nico wrote:
> On 2016/01/05 16:26:10, Robert Sesek wrote:
> > On 2016/01/05 16:18:30, Nico wrote:
> > > ?
> > 
> > It's for testing a static_assert.
> 
> This should have a comment saying so, else it's be deleted by the next person
> touching this file. Or it should be in a .nc file.

Added a comment in https://codereview.chromium.org/1563463002 since .nc tests
don't seem to work on my machine.

[Mark Mentovai, 2016-01-06 16:22:58]

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_launch_...
File base/mac/scoped_launch_data.h (right):

https://codereview.chromium.org/1551943002/diff/40001/base/mac/scoped_launch_...
base/mac/scoped_launch_data.h:19: static void Free(launch_data_t ldt) {
launch_data_free(ldt); }
I had to ignore -Wdeprecated-declarations to use launch_data_free in
mini_chromium:

https://codereview.chromium.org/1565803002/diff/1/base/mac/scoped_launch_data...

because you get the system SDK and deployment target by default with
mini_chromium.