Chromium glue for Preemptive cursors and passing keys from renderer to browser

Chromium glue for Preemptive cursors and passing keys from renderer to browser

This will allow indexing to happen renderer-side

BUG=134493
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=151047

[alecflett, 2012-08-08 22:15:47]

Here's the chrome side - just boilerplate IPC, with one note about setIndexKeys
being async. This totally could be async, since there are no out parameters. Do
either of you have any thoughts about that? I'm thinking I should leave it sync
to be consistent with the rest of the API and wait for idb.next...

[alecflett, 2012-08-08 22:15:48]

Here's the chrome side - just boilerplate IPC, with one note about setIndexKeys
being async. This totally could be async, since there are no out parameters. Do
either of you have any thoughts about that? I'm thinking I should leave it sync
to be consistent with the rest of the API and wait for idb.next...

[dgrogan, 2012-08-08 23:02:49]

lgtm

I haven't been following this effort too closely, you should probably let Josh
take a look before submitting.

https://chromiumcodereview.appspot.com/10830228/diff/1/content/common/indexed...
File content/common/indexed_db/indexed_db_dispatcher.cc (right):

https://chromiumcodereview.appspot.com/10830228/diff/1/content/common/indexed...
content/common/indexed_db/indexed_db_dispatcher.cc:459: std::vector<string16>
index_names(web_index_names.size());
It looks like you could move this code to the call site in
RendererWebIDBObjectStoreImpl because you're not using any of
IndexedDBDispatcher's member variables.

I remember you asking about this a while ago.  Instead of the rambling answer
that I gave, I should have said "use the static IndexedDBDispatcher::Send if you
can, resort to dispatcher->RequestIDB... if you have to."

https://chromiumcodereview.appspot.com/10830228/diff/1/content/common/indexed...
File content/common/indexed_db/indexed_db_messages.h (right):

https://chromiumcodereview.appspot.com/10830228/diff/1/content/common/indexed...
content/common/indexed_db/indexed_db_messages.h:464: // FIXME: should this be
async?
Would the only code difference be to remove _SYNC from this message declaration?
 If so, sounds free and we should do it.

[jsbell, 2012-08-08 23:34:08]

lgtm as well...

I think I'm coming around to the renderer-driven index population approach,
although it will make the methods (like openCursor) that both implement the IDB
API and are used for this purpose more cluttered.

https://chromiumcodereview.appspot.com/10830228/diff/1/content/common/indexed...
File content/common/indexed_db/indexed_db_dispatcher.h (right):

https://chromiumcodereview.appspot.com/10830228/diff/1/content/common/indexed...
content/common/indexed_db/indexed_db_dispatcher.h:207:
WebKit::WebIDBCursor::Direction direction,
Just a reminder to add the WebIDBCursor::Direction <-> IDBCursor::Direction
tests to WebKit's Source/WebKit/chromium/src/AssertMatchingEnums.cpp

[alecflett, 2012-08-09 20:54:18]

Issues addressed..landing now

http://codereview.chromium.org/10830228/diff/1/content/common/indexed_db/inde...
File content/common/indexed_db/indexed_db_dispatcher.cc (right):

http://codereview.chromium.org/10830228/diff/1/content/common/indexed_db/inde...
content/common/indexed_db/indexed_db_dispatcher.cc:459: std::vector<string16>
index_names(web_index_names.size());
On 2012/08/08 23:02:50, dgrogan wrote:
> It looks like you could move this code to the call site in
> RendererWebIDBObjectStoreImpl because you're not using any of
> IndexedDBDispatcher's member variables.
> 
> I remember you asking about this a while ago.  Instead of the rambling answer
> that I gave, I should have said "use the static IndexedDBDispatcher::Send if
you
> can, resort to dispatcher->RequestIDB... if you have to."

Done.

http://codereview.chromium.org/10830228/diff/1/content/common/indexed_db/inde...
File content/common/indexed_db/indexed_db_dispatcher.h (right):

http://codereview.chromium.org/10830228/diff/1/content/common/indexed_db/inde...
content/common/indexed_db/indexed_db_dispatcher.h:207:
WebKit::WebIDBCursor::Direction direction,
On 2012/08/08 23:34:08, jsbell wrote:
> Just a reminder to add the WebIDBCursor::Direction <-> IDBCursor::Direction
> tests to WebKit's Source/WebKit/chromium/src/AssertMatchingEnums.cpp

good point! https://bugs.webkit.org/show_bug.cgi?id=93618

http://codereview.chromium.org/10830228/diff/1/content/common/indexed_db/inde...
File content/common/indexed_db/indexed_db_messages.h (right):

http://codereview.chromium.org/10830228/diff/1/content/common/indexed_db/inde...
content/common/indexed_db/indexed_db_messages.h:464: // FIXME: should this be
async?
On 2012/08/08 23:02:50, dgrogan wrote:
> Would the only code difference be to remove _SYNC from this message
declaration?
>  If so, sounds free and we should do it.

Done.

[commit-bot: I haz the power, 2012-08-09 20:54:31]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/alecflett@chromium.org/10830228/10

[commit-bot: I haz the power, 2012-08-10 15:29:35]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/alecflett@chromium.org/10830228/10

[commit-bot: I haz the power, 2012-08-10 16:35:09]

Change committed as 151047