RESOLVED FIXED Bug 14953
Implement window.console in WebCore
https://bugs.webkit.org/show_bug.cgi?id=14953
Summary Implement window.console in WebCore
David Kilzer (:ddkilzer)
Reported 2007-08-12 16:28:07 PDT
The first step to supporting the Firebug Console API (see Bug 14354) is to implement window.console in WebCore. (Currently, both Safari for Windows and Safari for Mac OS X implement it within the browser.) Will attach a patch shortly.
Attachments
Patch v1 (20.97 KB, patch)
2007-08-12 16:42 PDT, David Kilzer (:ddkilzer)
ddkilzer: review-
Updated patch [WIP] (22.32 KB, patch)
2007-08-12 17:58 PDT, Sam Weinig
no flags
Even more updated [Still WIP] (23.79 KB, patch)
2007-08-12 18:56 PDT, Sam Weinig
no flags
patch (30.37 KB, patch)
2007-10-27 19:05 PDT, Sam Weinig
aroben: review+
David Kilzer (:ddkilzer)
Comment 1 2007-08-12 16:42:40 PDT
Created attachment 15943 [details] Patch v1 Initial implementation of window.console in WebCore.
David Kilzer (:ddkilzer)
Comment 2 2007-08-12 17:08:10 PDT
Comment on attachment 15943 [details] Patch v1 Maciej says: We've used shadow/shadowing to refer specifically to overriding a built-in Window property with a var declaration, I think a better name for this concept would be "Replaceable".
David Kilzer (:ddkilzer)
Comment 3 2007-08-12 17:31:29 PDT
Comment on attachment 15943 [details] Patch v1 Wrong license applied to source files as well. Should be BSD-style, not LGPL.
Sam Weinig
Comment 4 2007-08-12 17:58:37 PDT
Created attachment 15944 [details] Updated patch [WIP] Adding an updated patch that is not quite ready for review. This moves Console.{h,cpp,idl} to WebCore/page/ with the rest of the window objects, fixes the licenses, fixes some style issues and changes the extended attribute to Replaceable.
David Kilzer (:ddkilzer)
Comment 5 2007-08-12 18:43:45 PDT
(In reply to comment #4) > Created an attachment (id=15944) [edit] > Updated patch [WIP] WebCore.pro and WebCore.vcproj changes were lost from previous patch.
Sam Weinig
Comment 6 2007-08-12 18:56:36 PDT
Created attachment 15946 [details] Even more updated [Still WIP] Adds WebCore.pro and WebCore.vcproj back
David Kilzer (:ddkilzer)
Comment 7 2007-10-13 02:50:50 PDT
Timothy Hatcher
Comment 8 2007-10-27 12:39:50 PDT
+ // FIXME: this should use LogMessageLevel, but must use ErrorMessageLevel for now to forward message to the console. etc. We should pass the right message level to Crome and let Crome's addMessageToConsole do the right thing when calling the client. Passing the wrong message level is gimping the new Web Inspector console.
Timothy Hatcher
Comment 9 2007-10-27 12:53:19 PDT
If you make this change: Index: page/Chrome.cpp =================================================================== --- page/Chrome.cpp (revision 27149) +++ page/Chrome.cpp (working copy) @@ -197,7 +197,7 @@ void Chrome::addMessageToConsole(MessageSource source, MessageLevel level, const String& message, unsigned lineNumber, const String& sourceID) { - if (source == JSMessageSource && level == ErrorMessageLevel) + if (source == JSMessageSource) m_client->addMessageToConsole(message, lineNumber, sourceID); if (InspectorController* inspector = m_page->inspectorController()) Then you can pass all the correct message levels.
Adam Roben (:aroben)
Comment 10 2007-10-27 16:00:29 PDT
(In reply to comment #8) > + // FIXME: this should use LogMessageLevel, but must use ErrorMessageLevel > for now to forward message to the console. > > etc. > > We should pass the right message level to Crome and let Crome's > addMessageToConsole do the right thing when calling the client. Passing the > wrong message level is gimping the new Web Inspector console. When I added the message level/source stuff I was being pretty paranoid about not changing the set of messages passed up to the ChromeClient. I think we can be less paranoid now and do as you suggest.
Sam Weinig
Comment 11 2007-10-27 19:05:04 PDT
Adam Roben (:aroben)
Comment 12 2007-10-27 19:19:54 PDT
Comment on attachment 16904 [details] patch It might be nice to add a test that actually called methods on the Console object. r=me
Sam Weinig
Comment 13 2007-10-27 19:25:36 PDT
Landed in r27161
Timothy Hatcher
Comment 14 2007-12-05 14:12:47 PST
*** Bug 14263 has been marked as a duplicate of this bug. ***
Note You need to log in before you can comment on or make changes to this bug.