Bug 46679 – Delete and forward delete shouldn't start autocorrection panel timer.

• Component: New Bugs
• Status: RESOLVED FIXED

[Jia Pu]

<rdar://problem/8463917>

Autocorrection panel should not show up when user is deleting or forward deleting.

[Jia Pu]

Created attachment 69004 [details]
Proposed patch (v1)

[WebKit Commit Bot]

Comment on attachment 69004 [details]
Proposed patch (v1)

Rejecting patch 69004 from commit-queue.

Failed to run "['./WebKitTools/Scripts/webkit-patch', '--status-host=queues.webkit.org', 'build', '--no-clean', '--no-update', '--build', '--build-style=both', '--quiet']" exit_code: 2
Building WebKit
Failed to run "['WebKitTools/Scripts/build-webkit', '--debug']" exit_code: 1

Full output: http://queues.webkit.org/results/4059172

[Adele Peterson]

Comment on attachment 69004 [details]
Proposed patch (v1)

I'm not sure why the cq failed- trying again...

[Eric Seidel (no email)]

I'm not sure either.  I suspect this is a bug from the CommitQueueTask rewrite.

I suspect we're not raising the right ScriptError, thus we're losing the log information?

[WebKit Commit Bot]

Comment on attachment 69004 [details]
Proposed patch (v1)

Rejecting patch 69004 from commit-queue.

Failed to run "['./WebKitTools/Scripts/webkit-patch', '--status-host=queues.webkit.org', 'build', '--no-clean', '--no-update', '--build', '--build-style=both', '--quiet']" exit_code: 2
Building WebKit
Failed to run "['WebKitTools/Scripts/build-webkit', '--debug']" exit_code: 1

Full output: http://queues.webkit.org/results/4111094

[Eric Seidel (no email)]

This patch doesn't seem to build correctly on the mac.  However the commit-queue's rejection path is not correctly spitting out the build log.  Investigating.

[Eric Seidel (no email)]

http://trac.webkit.org/browser/trunk/WebKitTools/Scripts/webkitpy/tool/bot/commitqueuetask.py#L142

is the path it should be hitting.  It's possible somehow the ScriptError which was saved is losing its log?

[Adele Peterson]

I'm building this locally, and will land manually eventually.

[Adam Barth]

> I suspect we're not raising the right ScriptError, thus we're losing the log information?

The problem is that the bug comment is now coming from the master process, but the child process never outputs any build spew.  We probably want the child to print some of the build spew on failure so the parent can see it.

[Adele Peterson]

TypingCommand.cpp:286: warning: unused parameter 'commandType'

This is the warning that I'm getting.

[Adele Peterson]

Comment on attachment 69004 [details]
Proposed patch (v1)

r- so Jia can fix the warning.

[Jia Pu]

Created attachment 69099 [details]
Proposed patch (v2)

Fixed a build failure on non-Mac OSX platforms. Not sure if this is cleanest way to fix it. Any suggestion from reviewers is welcome.

[Adele Peterson]

Comment on attachment 69099 [details]
Proposed patch (v2)

I'm not sure about this.  I was seeing the warning on 10.6.4.

Comment on attachment 69099 [details]
Proposed patch (v2)

View in context: https://bugs.webkit.org/attachment.cgi?id=69099&action=review

> WebCore/editing/TypingCommand.cpp:321
> +#if PLATFORM(MAC) && !defined(BUILDING_ON_TIGER) && !defined(BUILDING_ON_LEOPARD)
>  void TypingCommand::typingAddedToOpenCommand(ETypingCommand commandTypeForAddedTyping)
> +#else
> +void TypingCommand::typingAddedToOpenCommand(ETypingCommand)
> +#endif

The complaint was about markMisspellingsAfterTyping() having an unused (in some configurations) parameter, not about this method, so I’m not sure why you’re changing this one. I think it will break all builds.

The way to fix the warning about markMisspellingsAfterTyping() is to add
UNUSED_PARAM(commandType)
for the configurations that don’t use commandType.

You’ll need to make sure that you’re including <wtf/UnusedParam.h>—but I think you have it by way of including Editor.h.

[Early Warning System Bot]

Attachment 69099 [details] did not build on qt:
Build output: http://queues.webkit.org/results/4211001

[Jia Pu]

(In reply to comment #14)
> (From update of attachment 69099 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=69099&action=review
> 
> > WebCore/editing/TypingCommand.cpp:321
> > +#if PLATFORM(MAC) && !defined(BUILDING_ON_TIGER) && !defined(BUILDING_ON_LEOPARD)
> >  void TypingCommand::typingAddedToOpenCommand(ETypingCommand commandTypeForAddedTyping)
> > +#else
> > +void TypingCommand::typingAddedToOpenCommand(ETypingCommand)
> > +#endif
> 
> The complaint was about markMisspellingsAfterTyping() having an unused (in some configurations) parameter, not about this method, so I’m not sure why you’re changing this one. I think it will break all builds.
> 
Sorry, totally messed up this one. Let me generate another patch, and test building it on both platforms

[Jia Pu]

Created attachment 69127 [details]
Proposed patch (v3)

[WebKit Commit Bot]

Comment on attachment 69127 [details]
Proposed patch (v3)

Clearing flags on attachment: 69127

Committed r68613: <http://trac.webkit.org/changeset/68613>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.