(trybot test) base: Make ScopedPtrMap use DISALLOW_COPY_AND_ASSIGN

(trybot test) base: Make ScopedPtrMap use DISALLOW_COPY_AND_ASSIGN

Currently the class uses MOVE_ONLY_TYPE_WITH_MOVE_CONSTRUCTOR_FOR_CPP_03
which adds a .Pass() method and works for passing with Callback. But
there are no users of this passing ownership in Callback, so switch
to the general DISALLOW_COPY_AND_ASSIGN macro.

This unconvered a use of ScopedPtrMap::Pass() though, so I just changed
it to use a std::map instead of ScopedPtrMap.

R=Nico, zea@chromium.org
BUG=557422, 554291
patch from issue 1480773002 at patchset 40001 (http://crrev.com/1480773002#ps40001)

[vmpstr, 2015-11-30 23:15:46]

Description was changed from

==========
(trybot test) base: Make ScopedPtrMap use DISALLOW_COPY_AND_ASSIGN

Currently the class uses MOVE_ONLY_TYPE_WITH_MOVE_CONSTRUCTOR_FOR_CPP_03
which adds a .Pass() method and works for passing with Callback. But
there are no users of this passing ownership in Callback, so switch
to the general DISALLOW_COPY_AND_ASSIGN macro.

This unconvered a use of ScopedPtrMap::Pass() though, so I just changed
it to use a std::map instead of ScopedPtrMap.

R=Nico, zea@chromium.org
BUG=557422, 554291

patch from issue 1480773002 at patchset 40001
(http://crrev.com/1480773002#ps40001)
==========

to

==========
(trybot test) base: Make ScopedPtrMap use DISALLOW_COPY_AND_ASSIGN

Currently the class uses MOVE_ONLY_TYPE_WITH_MOVE_CONSTRUCTOR_FOR_CPP_03
which adds a .Pass() method and works for passing with Callback. But
there are no users of this passing ownership in Callback, so switch
to the general DISALLOW_COPY_AND_ASSIGN macro.

This unconvered a use of ScopedPtrMap::Pass() though, so I just changed
it to use a std::map instead of ScopedPtrMap.

R=Nico, zea@chromium.org
BUG=557422, 554291

patch from issue 1480773002 at patchset 40001
(http://crrev.com/1480773002#ps40001)
==========

[vmpstr, 2015-11-30 23:15:46]

vmpstr@chromium.org changed reviewers:
- thakis@chromium.org, zea@chromium.org

[danakj, 2015-12-01 00:42:34]

danakj@chromium.org changed reviewers:
+ danakj@chromium.org, thakis@chromium.org, zea@chromium.org

[danakj, 2015-12-01 00:42:34]

Would you like to rebase this on my CL then to convert the type? Thanks for
figuring this out!

[Nico, 2015-12-01 01:02:16]

thanks!

https://codereview.chromium.org/1490613002/diff/1/sync/engine/commit_processo...
File sync/engine/commit_processor.cc (right):

https://codereview.chromium.org/1490613002/diff/1/sync/engine/commit_processo...
sync/engine/commit_processor.cc:44: (*contributions)[it.Get()] =
std::move(contribution);
Should there be a comment here saying that this uses foo[] instead of
foo.insert() to work around an msvs2013 bug?

[danakj, 2015-12-01 01:03:12]

https://codereview.chromium.org/1490613002/diff/1/sync/engine/commit_processo...
File sync/engine/commit_processor.cc (right):

https://codereview.chromium.org/1490613002/diff/1/sync/engine/commit_processo...
sync/engine/commit_processor.cc:44: (*contributions)[it.Get()] =
std::move(contribution);
On 2015/12/01 01:02:16, Nico wrote:
> Should there be a comment here saying that this uses foo[] instead of
> foo.insert() to work around an msvs2013 bug?

Oh, or maybe we don't need to do this and can insert(std::make_pair(...)) cuz
this didn't fix anything anyways.

[vmpstr, 2015-12-01 01:03:58]

https://codereview.chromium.org/1490613002/diff/1/sync/engine/commit_processo...
File sync/engine/commit_processor.cc (right):

https://codereview.chromium.org/1490613002/diff/1/sync/engine/commit_processo...
sync/engine/commit_processor.cc:44: (*contributions)[it.Get()] =
std::move(contribution);
On 2015/12/01 01:02:16, Nico wrote:
> Should there be a comment here saying that this uses foo[] instead of
> foo.insert() to work around an msvs2013 bug?

Oh this was just me applying the latest patchset, I think a pair would work
here. I'll create a different review once the other patch lands.

I'm closing this for now.

[Nico, 2015-12-01 01:04:19]

Ah, just saw the discussion on the other CL. Maybe the CL description could
describe the actual workaround then :-)