Bunch of match preview tweaks:

Bunch of match preview tweaks:
. Makes MatchPreview owned by Browser rather than each TabContents.
. Makes MatchPreview dismiss when the omnibox closes.
. Supports the ability to send script to the page rather than
  reloading on every keystroke.
. Supports receiving results from the page that drives the suggest
  text in the omnbox.

BUG=54833
TEST=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=59428

[Jay Civelli, 2010-09-14 01:10:30]

http://codereview.chromium.org/3332022/diff/9001/10007
File chrome/browser/browser.h (right):

http://codereview.chromium.org/3332022/diff/9001/10007#newcode40
chrome/browser/browser.h:40: class MatchPreview;
You don't need this if you keep the include for match_preview.h.
If you decide to make MatchPreviewDelegate its own file, then you can get rid of
the include instead.

http://codereview.chromium.org/3332022/diff/9001/10007#newcode171
chrome/browser/browser.h:171: MatchPreview* match_preview() { return
match_preview_.get(); }
Make it const.

http://codereview.chromium.org/3332022/diff/9001/10011
File chrome/browser/tab_contents/match_preview.cc (right):

http://codereview.chromium.org/3332022/diff/9001/10011#newcode62
chrome/browser/tab_contents/match_preview.cc:62: explicit
FrameLoadObserver(MatchPreview* match_preview,
Remove explicit.

http://codereview.chromium.org/3332022/diff/9001/10011#newcode454
chrome/browser/tab_contents/match_preview.cc:454: url =
GURL("http://www/~sky/no_crawl/user_inputted2.html");
Is it supposed to be removed?

http://codereview.chromium.org/3332022/diff/9001/10012
File chrome/browser/tab_contents/match_preview.h (right):

http://codereview.chromium.org/3332022/diff/9001/10012#newcode20
chrome/browser/tab_contents/match_preview.h:20: class MatchPreviewDelegate {
To limit headers explosions, as Eliot suggested in an email may be this should
be in its own file?

http://codereview.chromium.org/3332022/diff/9001/10020
File chrome/browser/views/frame/browser_view_layout.cc (right):

http://codereview.chromium.org/3332022/diff/9001/10020#newcode214
chrome/browser/views/frame/browser_view_layout.cc:214: contents_container_ =
static_cast<ContentsContainer*>(
May be we could use some magic view IDs to find the sidebar_split children more
reliably?

http://codereview.chromium.org/3332022/diff/9001/10022
File chrome/browser/views/frame/contents_container.cc (right):

http://codereview.chromium.org/3332022/diff/9001/10022#newcode31
chrome/browser/views/frame/contents_container.cc:31: ~TearWindow() {
Make virtual.

http://codereview.chromium.org/3332022/diff/9001/10024
File chrome/browser/views/location_bar/location_bar_view.cc (right):

http://codereview.chromium.org/3332022/diff/9001/10024#newcode1183
chrome/browser/views/location_bar/location_bar_view.cc:1183: //
RenderWidgetHostView (or one of it's children).
Nit: it's -> its

http://codereview.chromium.org/3332022/diff/9001/10024#newcode1187
chrome/browser/views/location_bar/location_bar_view.cc:1187:
match_preview->preview_contents()->GetRenderWidgetHostView()) {
Nit: you could return early.

[sky, 2010-09-14 04:03:05]

Almost all addressed. New snapshot uploaded.

  -Scott

http://codereview.chromium.org/3332022/diff/9001/10020
File chrome/browser/views/frame/browser_view_layout.cc (right):

http://codereview.chromium.org/3332022/diff/9001/10020#newcode214
chrome/browser/views/frame/browser_view_layout.cc:214: contents_container_ =
static_cast<ContentsContainer*>(
On 2010/09/14 01:10:31, Jay Civelli wrote:
> May be we could use some magic view IDs to find the sidebar_split children
more
> reliably?

Or just use real methods. I'll fix this later though.

[Jay Civelli, 2010-09-14 16:25:02]

LGTM