Merge lp://staging/~abreu-alexandre/oxide/donottrack into lp://staging/~oxide-developers/oxide/oxide.trunk

Proposed by Alexandre Abreu
Status: Merged
Merged at revision: 1159
Proposed branch: lp://staging/~abreu-alexandre/oxide/donottrack
Merge into: lp://staging/~oxide-developers/oxide/oxide.trunk
Diff against target: 438 lines (+212/-1) (has conflicts)
13 files modified
qt/core/browser/oxide_qt_web_context.cc (+24/-0)
qt/core/browser/oxide_qt_web_context.h (+3/-0)
qt/core/glue/oxide_qt_web_context_proxy.h (+3/-0)
qt/quick/api/oxideqquickwebcontext.cc (+19/-0)
qt/quick/api/oxideqquickwebcontext_p.h (+6/-0)
qt/tests/qmltests/api/tst_WebContext_doNotTrack.py (+18/-0)
qt/tests/qmltests/api/tst_WebContext_doNotTrack.qml (+74/-0)
shared/browser/oxide_browser_context.cc (+33/-1)
shared/browser/oxide_browser_context.h (+5/-0)
shared/browser/oxide_browser_context_observer.h (+2/-0)
shared/browser/oxide_network_delegate.cc (+10/-0)
shared/browser/oxide_web_view_contents_helper.cc (+14/-0)
shared/browser/oxide_web_view_contents_helper.h (+1/-0)
Text conflict in qt/core/browser/oxide_qt_web_context.cc
To merge this branch: bzr merge lp://staging/~abreu-alexandre/oxide/donottrack
Reviewer Review Type Date Requested Status
Chris Coulson Approve
Review via email: mp+261766@code.staging.launchpad.net

Commit message

Handle do not track

Description of the change

Handle do not track

To post a comment you must log in.
1126. By Chris Coulson

Bump Chromium rev to 45.0.2427.7

1127. By Chris Coulson

Persist media device permissions for the life of WebContext

1128. By Chris Coulson

Make ContentBrowserClient / ContentRendererClient memebers of ContentMainDelegate like ContentClient already is

1129. By Chris Coulson

Pass the application locale to ui::ResourceBundle::InitSharedInstanceWithLocale. Because we don't build with glib support, it's unable to detect the system locale and depends on a preferred one being passed

1130. By Chris Coulson

Convert UserScriptMaster in to a KeyedService

1131. By Chris Coulson

Create a single TestWebContext in tst_WebView_save_restore_state.qml

1132. By Chris Coulson

Don't allow QObjects to be passed to WebContextDelegateWorker

1133. By Chris Coulson

Add justifications around less-obvious uses of base::Unretained

1134. By Olivier Tilloy

Fix save/restore unit tests.

1135. By Chris Coulson

Bump Chromium rev to 45.0.2431.0

1136. By Chris Coulson

Fix an intermittent test failure

1137. By Chris Coulson

Ensure we have a WebScopedMicrotaskSuppression instance on the stack when we call in to V8

1138. By Chris Coulson

Add a basic editorconfig file - you can get editorconfig plugins for various editors from http://editorconfig.org/. This avoids having to turn on support for modelines, which is disabled in some editors (eg, vim) for security reasons

Revision history for this message
Chris Coulson (chrisccoulson) wrote :

Thanks, I've left some comments inline.

Also, the test is missing a few bits:

- It should verify that the default is off.
- It should check that the notify signal is emitted when you toggle it.
- What happens to window.navigator.doNotTrack if you change the setting without reloading the page? Does it update automatically? If it does, then that should be tested.

review: Needs Fixing
1139. By Chris Coulson

Some updates for the content script messaging:
- The payload is no longer converted to a string for the wire - instead, we use base::Value directly.
- As a consequence, on the renderer side, we now use content::V8ValueConverter to convert the payload from v8::Value to base::Value. This means we can get rid of the hacky script we compile for each script context, just to call JSON.stringify.
- On the browser-side, we now convert from base::Value directly to QVariant using a custom converter. There is no longer a dependency on QJsonDocument.
- As a hardening measure, our converter on the browser side protects against excessive recursion that would otherwise exhaust stack space.
- The "args" parameter for script messages is deprecated - use "payload" instead. "args" never really made sense as it implies it's always a list, and is there as a consequence of this API being put together in haste to get unit tests working.
- Message payloads no longer have to be JSON arrays or objects - they can be any primitive that can be represented in JSON.
- The script message dispatcher on the browser side is no longer a BrowserMessageFilter running on the IO thread, but instead a WebContentsObserver. We can do that now since it has an API to expose the source RenderFrameHost.

1140. By Chris Coulson

Use the preferred allow() method for the geolocation permission tests (we still retain one test for the deprecated accept() method though)

1141. By Chris Coulson

Store TestApiHost in TestWebView, keyed by WebFrame

1142. By Chris Coulson

Add TestUtils.waitFor, so that we can stop duplicating the same function for tests that don't use TestWebView

1143. By Chris Coulson

Bump depot_tools to c1ae89ecd635abfea2d94e5b49c7d92f49f28f22

1144. By Chris Coulson

Improve script messaging tests

1145. By Chris Coulson

LP: #1466208

1146. By Chris Coulson

Remove a spurious console.log

1147. By Chris Coulson

Correctly detect when we are running on Mir/Mesa

1148. By Chris Coulson

Assume that the display is already initialized

1149. By Chris Coulson

Bump Chromium rev to 45.0.2438.3

1150. By Chris Coulson

Add WebContext.userAgentOverrides API, which is a replacement for WebContext.userAgentOverrideDelegate. The new API allows applications to provide a list of regular expressions and user agent overrides as opposed to providing a JS worker that we call in to to ask for a user agent string. This allows us to stop relying on synchronous IPC to the browser

1151. By Chris Coulson

We can actually lazyily create UserAgentSettings. RenderProcessInitializer will create it when the first render process starts anyway

1152. By Alexandre Abreu

Fixes

Revision history for this message
Alexandre Abreu (abreu-alexandre) wrote :

All comments have been addressed

Revision history for this message
Chris Coulson (chrisccoulson) :
review: Approve

Preview Diff

[H/L] Next/Prev Comment, [J/K] Next/Prev File, [N/P] Next/Prev Hunk
The diff is not available at this time. You can reload the page or download it.

Subscribers

People subscribed via source and target branches