Use the same code path for all fullscreen transitions so that Chrome UI updates
correctly. All user-initiated fullscreen transitions now result in
CefWindowDelegate::OnWindowFullscreenTransition callbacks.
This avoids a situation where misbehaving clients may cause the
application to continue running indefinitely by posting new UI
thread tasks after calling CefQuitMessageLoop.
A modal dialog is a child CefWindow that implements some special behaviors
relative to a parent CefWindow. Like any CefWindow it can be framed with
titlebar or frameless, and optionally contain draggable regions (subject to
platform limitations described below). Modal dialogs are shown centered on
the parent window (inside a single display) and always stay on top of the
parent window in z-order. Sizing behavior and available window buttons are
controlled via the usual CefWindowDelegate callbacks. For example, the dialog
can have a preferred size with resize, minimize and maximize disabled (via
GetPreferredSize, CanResize, CanMinimize and CanMaximize respectively).
This change adds support for two modality modes. With window modality all
controls in the parent window are disabled. With browser modality only the
browser view in the parent window is disabled.
Both modality modes require that a valid parent window be returned via
GetParentWindow. For window modality return true from IsWindowModalDialog
and call CefWindow::Show. For browser modality return false from
IsWindowModalDialog (the default value) and call
CefWindow::ShowAsBrowserModalDialog with a reference to the parent window's
browser view.
Window modal dialog behavior depends on the platform. On Windows and
Linux these dialogs have a titlebar and can be moved independent of the
parent window. On macOS these dialogs do not have a titlebar, move with
the parent window, and do not support draggable regions (because they are
implemented using sheets). On Linux disabling the parent window controls
requires a window manager the supports _NET_WM_STATE_MODAL.
Browser modal dialog behavior is similar on all platforms. The dialog will
be automatically sized and positioned relative to the parent window's
browser view. Closing the parent window or navigating the parent browser
view will dismiss the dialog. The dialog can also be moved independent of
the parent window though it will be recentered when the parent window
itself is resized or redisplayed. On MacOS the dialog will move along with
the parent window while on Windows and Linux the parent window can be moved
independently.
To test: Use the Tests > Dialog Window menu option in cefclient with Views
enabled (`--use-views` or `--enable-chrome-runtime` command-line flag).
Browser modal dialog is the default behavior. For window modal dialog add
the `--use-window-modal-dialog` command-line flag.
This change replaces existing CefSettings.user_data_path usage with
CefSettings.root_cache_path for better alignment with the Chrome runtime.
All files should now be written inside the root_cache_path directory.
This adds support for the three-finger-swipe navigation gesture with
Alloy/Views. The default implementation matches the Chrome runtime
and navigates the browser back/forward. We also add an Alloy/Views-
specific client callback in CefBrowserViewDelegate for optional
custom handling of the gesture event.
Use pointers to incomplete struct types. If an argument of a not suitable
type is passed to a function of any string collection,
C compiler will warn:
warning: passing argument 1 of <func> from incompatible pointer type
C++ compiler will raise an error, for example:
error: cannot convert 'cef_string_list_t' ... to 'cef_string_multimap_t'
With the previously used `void*` pointer types, such errors in a code were not
diagnosed.
Initialize the Chrome color mixer so that ColorProvider returns the expected
semi-transparent color for overlay background in video_overlay_window_views.cc
instead of the default opaque red color (kPlaceholderColor).
This is intended for usage with frameless windows that show the standard window
buttons, where resizing the title bar height changes the button offset. Returning a
different value from CefWindowDelegate::GetTitlebarHeight and forcing a resize of
the NSWindow's theme frame (see ViewsWindow::NudgeWindow) will update the
title bar height.
To test:
1. Run `cefclient --use-views --hide-frame --show-window-buttons --url=http://tests/window`
2. Enter a new value for title bar height and click the "Set Titlebar Height" button
This change adds new CefCommandHandler callbacks for optionally hiding
specific Chrome toolbar icons, buttons and app menu items.
To test: Run `cefclient --enable-chrome-runtime --filter-chrome-commands`.
Most icons, buttons and app/context menu items will be hidden.
The CefFrame::CreateURLRequest method is no longer supported in the renderer
process. Usage of this method was already limited to same-origin requests due
to renderer process CORS restrictions, and the underlying Blink API has now
been removed in https://crbug.com/1413912 (M112+).
Existing alternatives include CefURLRequest usage in the browser process, or
JavaScript XMLHttpRequest/fetch API usage in the renderer process.
Frameless windows now display as expected. Default traffic light buttons can
optionally be shown at configurable vertical position. Layout respects text
direction.
Chrome is always loading the primary user profile by default, so with this
change the CEF behavior more accurately reflects reality. Incognito contexts
can still be created explicitly via CefRequestContext::CreateContext.
Prior to this change, the default for the global context was an Incognito
profile based on the primary user profile. That caused request interception
to be bypassed in WillCreateURLLoaderFactory for profiles associated with
the "New window" and "New incognito window" commands. Those profiles, while
also being (or based on) the primary user profile, did not match the specific
Incognito profile assigned to the default global context.
After this change, the "New window" and "New incognito window" commands will
match the default global context when executed on a browser that was created
using the primary user profile.
Note: In current master (based on M111), document PiP partially works when
run with the `--enable-features=DocumentPictureInPictureAPI` command-line
flag. However, the document PiP implementation at this Chromium version is
missing fixes that have already been cherry-picked to the 5563 release
branch. Those fixes will only be available in master after the next Chromium
update (to M112).
Marshall Greenblatt
Julien Bouix
Chris Frank
Nik Pavlov
Takashi Hashida
Sergey Markelov