| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
native_theme to native_theme_win since its Windows-specific.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/259047
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28300 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/271001
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28292 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- in TabbedPaneView, the focus traversable were not properly hooked-up
- the NativeTextField was focusing the wrapper view instead of the actual
Textfield view.
Also in the FocusTraversal unit-test I had to select one of the radio-button that
the test uses, as the behavior of focusing a disabled button is different on
Windows (it forces any non disabled radio-button to become selected) than on Gtk.
BUG=None
TEST=Run the unit tests.
Review URL: http://codereview.chromium.org/242163
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28263 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
me know. Here's the warning I was getting without this:
dereferencing type-punned pointer will break strict-aliasing rules
BUG=none
TEST=none
TBR=derat
Review URL: http://codereview.chromium.org/261005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28233 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
avoid the focus problems.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/246099
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28228 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
turning off border work.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/256080
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28172 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
WM_KEYDOWN that resulted in an accelerator, as they would cause orphaned
KeyUp events to be sent to views.
This is Windows only for now, still needs to be ported to Linux.
BUG=23383
TEST=Run unit-tests. Ensure accelerators work as expected in Chrome.
Review URL: http://codereview.chromium.org/246074
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28096 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=none
TEST=successful builds
Review URL: http://codereview.chromium.org/256059
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28089 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
It seems the added subclassing is causing crashers in the field.
BUG=23861
TEST=None
TBR=beng
Review URL: http://codereview.chromium.org/259052
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28061 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
ability to continually paint. I've tweaked it so that we only schedule
a paint if the paint is happening from Gtk and not views.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/251084
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27998 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TEST=none
BUG=none
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27939 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
clipboard_util to clipboard_util_win since it's Windows-only. This patch makes
test_shell depend on app as well. There should be no logic change.
TEST=none
BUG=none
Review URL: http://codereview.chromium.org/260003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27937 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
into the gfx namespace.
Combine the PNGEncoder and PNGDecoder. There were separate when we had
different executables for the browser and renderer, and linked the encoder only
in one of them (which saved us some space used by libpng). This hasn't been the
case for years, so combining them (again) makes sense.
TEST=none
BUG=none
Review URL: http://codereview.chromium.org/243076
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27930 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
window that extends below a browser action button when the browser is clicked. When it loses focus, it is automatically dismissed.
BUG=23596
TEST=none
Review URL: http://codereview.chromium.org/258011
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27889 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
painted properly. This is the case for animations, such as the dangerous download item animation.
(For some unclear reasons this would happen only in aero and only in a popup browser frames.)
Explicitly asking for a repaint when setting the clip fixes it.
BUG=23087
TEST=Start a dangerous download from a popup (sorry, I don't have any public URL to repro). The dangerous download should display properly in the download shelf.
Review URL: http://codereview.chromium.org/259005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27858 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It was previously breaking the interactive UI tests.
That was because I changed the button base class to be focusable by default which led to many button beings added to the tab traversal order.
I now explicitly make these buttons non focusable.
Also because of the way we clear the native focus on Gtk (by setting focus to NULL on the top level window), views now default to clearing the focus when focused (so that views non backed by a native window don't leave the native focus on the previously native window).
On Windows we used to focus the WidgetWin's HWND, and now we focus the top-level HWND. The WidgetWin class had to be changed to forward the key events to the root view that contains the focused view because of that. (otherwise it would send it to the top-level root-view, and if the focused view was in a child root-view it would not recieve the keyboard messages).
Note that the mouse wheel does not need that, as the mouse-wheel Windows messages are sent to the HWND under the cursor.
See:
http://codereview.chromium.org/246030/show
TEST=Run the tests
BUG=NONE
Review URL: http://codereview.chromium.org/255008
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27723 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
paint we would expand the region Gtk wants to paint to include the
region views wants to paint. This is problematic if the views region
isn't contained in the gtk region. In such a case any gtk widgets
without windows aren't painted correctly. The correct thing to do is
schedule the paint through gtk so that the child widgets are
painted. Additionally I'm making WidgetGtk::PaintNow actually paint
now, which helps avoid this situation.
BUG=23430
TEST=see bug
Review URL: http://codereview.chromium.org/246047
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27648 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The problem was caused by the way we were setting the focus,
using grab_focus which does not work whenthe screen is locked.
We now explicitly send a signal.
Also make sure to give a good parent when building a child
GtkWidget, as using the WidgetGtk native window would cause
a Gtk error.
BUG=23394
TEST=Run the unit-tests.
R=oshima
Review URL: http://codereview.chromium.org/242071
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27629 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
investigate.
BUG=23394
TBR=huanr
Review URL: http://codereview.chromium.org/251039
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27570 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Refactoring some of the NativeViewHost and NativeControl focus management so their consumers don't have to explicitly set the focused view.
See original review:
http://codereview.chromium.org/235011/show
BUG=None
TEST=Run all tests. Make sure focus is stored/restored properly in Chrome.
TBR=ben
Review URL: http://codereview.chromium.org/246032
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27563 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/242045
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27537 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/242051
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27527 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
after review)
See http://codereview.chromium.org/222030 for the original review
BUG=none
TEST=compiled & tested chrome target on win/linux, view_examples target on linux.
Review URL: http://codereview.chromium.org/242036
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27513 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TBR=oshima
Review URL: http://codereview.chromium.org/243034
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27512 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
http://codereview.chromium.org/251001/show
This CL hooks the focus traversal for the TabbedPaneView on Linux with toolkit views.
It also has few focus related fixed (focus with buttons, clearing focus effectively in the focus manager).
BUG=None
TEST=Run the view_examples app. Press tab to move the focus. Try the different tabs.
Review URL: http://codereview.chromium.org/246030
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27504 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=23071
TEST=none
Review URL: http://codereview.chromium.org/251024
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27449 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
http://codereview.chromium.org/248010/show
Changing the KeyboardEvent to use a KeyboardCode instead of a w_char. Led to several places where I had to switch from VK_ to VKEY_.
Also cleaned-up the table view OnKeyDown method. Since TableView is a NativeControl it can use the NativeControl::OnKeyDown directly.
BUG=None
TEST=Make sure short-cuts works as expected, especially in the omnibox.
Review URL: http://codereview.chromium.org/251020
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27444 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
widget_win.cc. This significantly reduces build times when modifying
the code of these event handlers during debugging.
BUG=http://code.google.com/p/chromium/issues/detail?id=23274
TEST=none
Review URL: http://codereview.chromium.org/254003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27435 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TBR=sky
Review URL: http://codereview.chromium.org/254005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27416 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
popups/app windows.
Rather than carrying state in a force_native_frame_ member it uses a virtual method AlwaysUseNativeFrame analogous to AlwaysUseCustomFrame. This makes me a little happier.
BUG=none
TEST=On a vista capable system, test that when a theme is installed, popups and app frames are rendered with the native frame. Test that constrained windows (e.g. HTTP basic auth) are rendered with the custom frame. Without a theme installed, test that all windows have a native frame except constrained windows. With Vista Basic system setting, verify that all windows have a custom frame, including constrained windows.
Review URL: http://codereview.chromium.org/200146
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27413 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
several places where I had to switch from VK_ to VKEY_.
Also cleaned-up the table view OnKeyDown method. Since TableView is a NativeControl it can use the NativeControl::OnKeyDown directly.
BUG=None
TEST=Make sure short-cuts works as expected, especially in the omnibox.
Review URL: http://codereview.chromium.org/248010
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27412 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=23261
TEST=none
Review URL: http://codereview.chromium.org/249020
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27403 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
views. Converts from using notification server to observer as
notification service is chrome only.
Also changes the pointer type used by window_gtk to be a left arrow.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/245016
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27398 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
This reverts commit 27381 (build break).
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27384 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
versions of Chrome.
|owner_drawn| is currently set if the UI font family or the UI font size is overriden. However, this leads menu items to be invisible in the system menu. This was not the case in Chrome 2.0, but somehow the way 'owner_drawn' menu is drawn has changed and we have this regression.
As a work around for the branch (3.0 train left, but we can merge this to 3.0.1), this CL sets |owner_drawn| to false for the system menu even if the UI font or the UI font size is overriden.
BUG=18042 (http://crbug.com/18042 )
TEST=Run Chrome with '--lang=hi' command line flag on Windows. Make sure that the translated menu items in the system menu are visible (e.g. 'New Tab', 'Task Manager') although a bit too small to read while the font size in the page/wrench menu is reasonably large.
Review URL: http://codereview.chromium.org/200123
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27382 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
enabled ScrollView on linux.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/222030
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27381 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
for windows.
Please don't commit code that does not compile.
Review URL: http://codereview.chromium.org/244019
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27357 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
| |
Review URL: http://codereview.chromium.org/242030
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27356 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
| |
Review URL: http://codereview.chromium.org/248016
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27354 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
tbr:oshima@chromium.org
Review URL: http://codereview.chromium.org/242028
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27352 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
. I nuked NativeButton's minimum_size_ field as it's only used to
enforce a minimum size on windows. The code to adjust for minimum
size is now ifdef'd on windows.
. Changes GTK's button to set the font as appropriate.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/246007
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27288 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
interactive ui tests. Note that this was originally reviewed in
http://codereview.chromium.org/217022/show I originally Elliot
suggestion of replacing the usage of int for keycode with the
bae::Keycode type, but that led to the CL getting out of hands
(as this is used in many different places). So this is only the
patch set 1 of that CL, I'll replace the type in another CL]
Use windows keycodes under linux (and all non-windows platforms).
This fixes any place where we use a VKEY_* (RenderWidgetHost, for example)
under Linux, but breaks accelerators in TOOLKIT_VIEWS which relied on this
wrong behaviour.
Previously, keyboard_codes_linux defined all the VKEY_* constants as their
GDK_* counterparts, which is wrong since the VKEY_* are supposed to resolve
to windows key codes.
BUG=22551
TEST=Make sure accelerators still work as expected on Chrome Linux and
Chrome Linux with toolkit views. Test when the the accelerators with
the focus in the location bar and also with the focus on the page.
Review URL: http://codereview.chromium.org/235025
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27284 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
NativeControlGtk
* Layout if the visibility is changed after the native view is created.
(I also added DCHECK(GetWidget()) to make sure Wiget is available at this point)
NativeViewHostGtk
* decrement the native_view's ref count only if it still has a parent.
* create fixed in AddedToWidget if fixed_ does not exist. This can happen when
a view is removed, then added.
BUG=none
TEST=unit test for last 2 bugs is added to view_unittest.cc (ChangeVisibility).
Review URL: http://codereview.chromium.org/219035
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27282 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
| |
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27197 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
system WINDOW color for InfoBubbles, not white. Several places are made to explicitly use black labels where that's correct or respecting system colors is a non-trivial fix (bugs filed for most).
Also, apparently I didn't get all my cleanup changes in the last patch, oops.
BUG=92,21027
TEST=Set theme to high-contrast black on white, check that most pieces of chrome UI have readable text
Review URL: http://codereview.chromium.org/237005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27196 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/235012
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27157 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
change is the removal of some extra spacing Ben erroneously added to the link infobar a while back (clients already do their own spacing as needed, so this just added extra whitespace in mid-sentence).
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/223029
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27151 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
the expected behavoir is when focus is set to the toolbar (by hitting SHIFT+ALT+T) and toolbar buttons are traversed by right/left arrow keys, a bevel should appear on the button (HOT Image).
This patch fixes the initial hottracking issue where the first item is not hottracked.
BUG=10942
TEST=ALT+SHIFT+T and traverse with arrow keys, observe hot tracking.
Review URL: http://codereview.chromium.org/194037
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27145 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
This fixes leaks caused by not calling gtk_widget_destroy on
dialog windows when we exit.
While I'm here, fix a GList leak on Linux toolkit views.
Review URL: http://codereview.chromium.org/223015
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27130 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TBR=ben
Review URL: http://codereview.chromium.org/231022
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27115 0039d316-1c4b-4281-b951-d872f2087c98
|