In #20829, we fixed compositor recycling when switching between
BrowserViews, but it turns out that there is one additional case that we
need to handle. When we create a completely new BrowserView instance, it
starts of as visible (even when it hasn't been added to the window),
which means that it will need its own compositor instead of using the
recycled compositor.
To fix this, lets make BrowserViews hidden by default until they're
added to the window. See also #19988. This is a potentially breaking
change given that the initial value of `document.visibilityState` will
now be `hidden`, but given the experimental status of BrowserViews, I
think this is a fine change to make. The old behavior can be restored
with `webPreferences: { show: true }`.
Notes: Fix compositor recycling when creating new BrowserView
* fix: allow reading body from non-2xx responses in net.request (#21055)
* fix(urlrequest): allow non-2xx repsponse results
- closes#21046
* test(net): add test cases to verify non-2xx body
* test(session): update spec to match clientrequest behavior
* test(net): update test cases to match clientrequest behavior
* spec: clean up async net spec
* Update api-session-spec.js
* chore: fixup test as per original PR to master
This was a regression in #16125, which unintentionally put
`GlobalShortcutListener::RegisterAccelerator` into a
`#if defined(OS_MACOSX)` block.
Notes: Fix broken `globalShortcut.registerAll()` on Windows and Linux
* feat: add nativeTheme.themeSource to allow apps to override Chromiums theme choice (#19960)
* feat: add nativeTheme.shouldUseDarkColorsOverride to allow apps to override Chromiums theme choice
* spec: add tests for shouldUseDarkColorsOverride
* chore: add missing forward declarations
* refactor: rename overrideShouldUseDarkColors to themeSource
* chore: only run appLevelAppearance specs on Mojave and up
* chore: update patch with more info and no define
* Update spec-main/api-native-theme-spec.ts
Co-Authored-By: Jeremy Apthorp <jeremya@chromium.org>
* Update api-native-theme-spec.ts
* Update api-native-theme-spec.ts
* Update api-native-theme-spec.ts
* fix: don't expose nativeTheme in the renderer process (#20139)
Exposing these in the renderer didn't make sense as they weren't backed
by the same instance / value store. This API should be browser only
especially now that we have nativeTheme.themeSource. Exposing in
//common was a mistake from the beginning.
* fix: emit updated on NativeTheme on the UI thread to avoid DCHECK (#20137)
* fix: emit updated on NativeTheme on the UI thread to avoid DCHECK
* Update atom_api_native_theme.cc
* spec: wait a few ticks for async events to emit so that test events do not leak into each other
* chore: add SetGTKDarkThemeEnabled(enabled) internal helper to allow dynamic theme selection on linux (#19964)
This is just a after-creation setter for the `darkTheme` constructor option. This is delibrately
a method and not a property as there is no getter.
* spec: remove leftover .only
* feat: associate InProgressRequest with requests (#19648)
* feat: migrate webRequest module to NetworkService (Part 4) (#19679)
* chore: use gin in WebRequest
* Add stubs for APIs
* feat: migrate webRequest module to NetworkService (Part 5) (#19714)
* Pass WebRequest to ProxyingURLLoaderFactory
* Call WebRequestAPI in InProgressRequest
* Store the listeners
* Pass the request and response
* Add stub to handle the events
* Use extensions::WebRequestInfo
* Make sure webRequest is managed by Session
* chore: make creation of WebRequestNS more clear
* fix: check WebContents for service workers
* feat: migrate webRequest module to NetworkService (Part 6) (#19752)
* Implement OnBeforeSendHeaders
* Pass the request
* Handle simple listeners
* Handle response listeners
* Read responses from listener
* feat: migrate webRequest module to NetworkService (Part 7) (#19820)
* fix: gin treats Function as Dictionary when doing convertions
* fix: check if listener exists
* fix: listener callback should be executed in next tick
* feat: make InProgressRequest work
* test: re-enable protocol test that relies on webRequest
* chore: merge conditions
* feat: migrate webRequest module to NetworkService (Part 8) (#19841)
* fix: fill uploadData property
* fix: requestHeaders in onBeforeSendHeaders
* fix: responseHeaders in onHeadersReceived
* fix: header keys should not be lowercased
* fix: gin::Dictionary::Get succeeds even though key does not exist...
* fix: throw for invalid filters
* test: re-enable api-web-request-spec
* chore: do not use deprecated base::Value API
* feat: migrate webRequest module to NetworkService (Part 9) (#19976)
* no need to get WebContents for URLLoaderFactory
* consult embedder for network_factory created in net module
* set disable_web_security to false
* re-enable webRequest tests in net module
* feat: add new nativeTheme API
* chore: deprecate and clean up old systemPreferences theme APIs in favor of new nativeTheme module
* chore: clean up and deprecate things per feedback
* chore: add tests for deprecate and clean up invert impl
* build: when is a boolean not a boolean???
* restore stash
revert
some things work others dont
tracking area for rescue
manual popup
restore drag n drop
cleanup
* fix: make tray not block main process (#18880)
* fix: make tray not block main process
* make AtomMenuModel refcounted
* add support for ansi codes in title
add remove TODOs
* chore: use ScopedPumpMessagesInPrivateModes in tray (#18977)
* chore: use ScopedPumpMessagesInPrivateModes in tray
* revert refcounting of AtomMenuModel
* Prefer WeakPtr for posting tasks to handle unexpected destruction
* cleanup .h
* cleanup .mm
* add imports
add missing include
* fix: crash when tray popup called twice (#18999)
* remove highlightMode and TODOs
* remove unnecessary copy
Closes#11371.
Previously, we didn't consider the return value of the webRequest URLPattern mate converter, which meant that when the pattern wasn't correctly parsed owing to invalid filter specification users would not be made aware of that fact and would just think that the filtering itself had failed. This corrects that error by moving the business logic of url pattern parsing out of the converter and into the function itself so that granular and specific errors can be thrown.
There's also no real reason that i'm aware of not to allow wider breadth of filters by letting users use a wildcard for effective TLD, so I also overrode that (default for the 1-arg Parse is not to allow that).
Finally, I added some examples of url filter types for users to reference.
* test: move some BrowserWindow specs to the main process
* uncomment cross-site test
* move more tests
* re-enable, refactor and move visibilitychange specs
* move new-window event tests and re-enable them on mac
* move max/minimize event tests
* move modal tests
* move beginFrameSubscription tests
* move savePage test
* move BrowserWindow options argument is optional test
* move restore, unmaximize, fullscreen tests
* move parent window tests
* don't wait for show event on windows (#8664)
* add debugging logs to fullscreen tests
* more debugging on windows
* explicitly destroy browserviews to prevent crash during gc
* only await show on darwin
* more event timing fixes
* disable max/minimize event tests on linux, since they're broken on CI
