Always add flag "FLAG_ACTIVITY_NEW_TASK" when doing a Web Search

Always add flag "FLAG_ACTIVITY_NEW_TASK" when doing a Web Search

For Android WebView, conceptually it is better to create a new task to
perform a Web Search (which will be handled by Google Search App) than a
new activity under the same task.

BUG=

Committed: https://crrev.com/947e8d6611df52dec945fa1b7be3d8de267adf9e
Cr-Commit-Position: refs/heads/master@{#324967}

[hush (inactive), 2015-04-10 22:25:04]

hush@chromium.org changed reviewers:
+ dtrainor@chromium.org

[hush (inactive), 2015-04-10 22:25:04]

PTAL!
An example on Android:
if you go to gmail and long press on a word, then choose "Web Search", the
search result page will replace your email thread in the "Recent apps" screen.
Conceptually, you would want to switch between the email you were reading and
the google search page for a particular word.

[hush (inactive), 2015-04-10 22:29:00]

This was initially added in the following CL:
https://chromiumcodereview.appspot.com/23654021/diff/16001/content/public/and...

[David Trainor- moved to gerrit, 2015-04-11 00:41:01]

On 2015/04/10 22:29:00, hush wrote:
> This was initially added in the following CL:
>
https://chromiumcodereview.appspot.com/23654021/diff/16001/content/public/and...

It looks like I refactored that code out of another place actually.  I'm fine
with this as long as the Chrome behavior doesn't change.  I think it would
probably be safer though to leave this default behavior the same and add another
ContentViewClient method that AwContentViewClient overrides to specify whether
or not to spawn in a new task.  I know it's not a big deal, but ContentShell
would break, for example.

wdyt?

[hush (inactive), 2015-04-13 21:40:28]

On 2015/04/11 00:41:01, David Trainor wrote:
> On 2015/04/10 22:29:00, hush wrote:
> > This was initially added in the following CL:
> >
>
https://chromiumcodereview.appspot.com/23654021/diff/16001/content/public/and...
> 
> It looks like I refactored that code out of another place actually.  I'm fine
> with this as long as the Chrome behavior doesn't change.  I think it would
> probably be safer though to leave this default behavior the same and add
another
> ContentViewClient method that AwContentViewClient overrides to specify whether
> or not to spawn in a new task.  I know it's not a big deal, but ContentShell
> would break, for example.
> 
> wdyt?

Sure. I think that will do.
See PS2.
So now, both chrome and android webview will override performWebSearch. I still
can't remove the default logic in ContentViewCore.java though, because it will
be used by the various test shells.

[hush (inactive), 2015-04-13 21:40:50]

hush@chromium.org changed reviewers:
+ sgurun@chromium.org

[hush (inactive), 2015-04-13 21:40:51]

Selim, PTAL PS2

[David Trainor- moved to gerrit, 2015-04-13 23:38:25]

lgtm

[sgurun-gerrit only, 2015-04-13 23:54:15]

On 2015/04/13 23:38:25, David Trainor wrote:
> lgtm

lgtm

[hush (inactive), 2015-04-13 23:55:07]

The CQ bit was checked by hush@chromium.org

[commit-bot: I haz the power, 2015-04-13 23:55:51]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1070443003/20001

[commit-bot: I haz the power, 2015-04-14 00:46:43]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-04-14 00:48:08]

Patchset 2 (id:??) landed as
https://crrev.com/947e8d6611df52dec945fa1b7be3d8de267adf9e
Cr-Commit-Position: refs/heads/master@{#324967}