The [Callback] IDL extended attribute is not part of the Web IDL specification and should not be needed. The bindings generator should be smart enough to know by itself if a parameter's type is a callback interface. This will results in simpler IDL files and closer to the standard.
Created attachment 199853 [details] Patch
Comment on attachment 199853 [details] Patch Attachment 199853 [details] did not pass gtk-ews (gtk): Output: http://webkit-queues.appspot.com/results/133502
Comment on attachment 199853 [details] Patch Attachment 199853 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.appspot.com/results/133498
Comment on attachment 199853 [details] Patch Attachment 199853 [details] did not pass win-ews (win): Output: http://webkit-queues.appspot.com/results/188022
Comment on attachment 199853 [details] Patch Attachment 199853 [details] did not pass mac-ews (mac): Output: http://webkit-queues.appspot.com/results/133499
Created attachment 199856 [details] Patch
Comment on attachment 199856 [details] Patch Attachment 199856 [details] did not pass gtk-ews (gtk): Output: http://webkit-queues.appspot.com/results/14513
Created attachment 199863 [details] Patch
Created attachment 199865 [details] Patch
Created attachment 199866 [details] Patch
Comment on attachment 199866 [details] Patch Attachment 199866 [details] did not pass gtk-wk2-ews (gtk-wk2): Output: http://webkit-queues.appspot.com/results/43916
Comment on attachment 199866 [details] Patch Attachment 199866 [details] did not pass gtk-ews (gtk): Output: http://webkit-queues.appspot.com/results/13432
Created attachment 199873 [details] Patch Should fix GTK ews.
Comment on attachment 199873 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=199873&action=review > Source/WebCore/Modules/webdatabase/SQLTransaction.idl:37 > + in [Optional] SQLStatementCallback callback, Maybe you want to replace [Optional] with 'optional' when you have time. It's already done in Blink :) > Source/WebCore/bindings/scripts/CodeGenerator.pm:431 > + return 0 if $type eq "Array"; This wouldn't be needed since it is already checked in GetArrayType(). > Source/WebCore/bindings/scripts/CodeGenerator.pm:679 > + return 0 unless $object->IsRefPtrType($type); > + return 0 if $type eq "SerializedScriptValue"; Why do you want to write SerializedScriptValue here instead of in IsRefPtrType(). Actually I'm not sure if we should use IsRefPtrType() here. Maybe we should introduce IsWrapperType() for this use, just like V8 does.
Comment on attachment 199873 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=199873&action=review >> Source/WebCore/Modules/webdatabase/SQLTransaction.idl:37 >> + in [Optional] SQLStatementCallback callback, > > Maybe you want to replace [Optional] with 'optional' when you have time. It's already done in Blink :) On my TODO list :) >> Source/WebCore/bindings/scripts/CodeGenerator.pm:431 >> + return 0 if $type eq "Array"; > > This wouldn't be needed since it is already checked in GetArrayType(). ok >> Source/WebCore/bindings/scripts/CodeGenerator.pm:679 >> + return 0 if $type eq "SerializedScriptValue"; > > Why do you want to write SerializedScriptValue here instead of in IsRefPtrType(). > > Actually I'm not sure if we should use IsRefPtrType() here. Maybe we should introduce IsWrapperType() for this use, just like V8 does. Well, SerializedScriptValue is a RefPtr type so I cannot move it to IsRefPtrType(). Anyway, I think you are right, introducing an IsWrapperType() would probably be clearer and we would avoid editing existing functions.
Comment on attachment 199873 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=199873&action=review >>> Source/WebCore/bindings/scripts/CodeGenerator.pm:431 >>> + return 0 if $type eq "Array"; >> >> This wouldn't be needed since it is already checked in GetArrayType(). > > ok Actually GetArrayType() return the array subtype: return $1 if $type =~ /^([\w\d_\s]+)\[\]/; Therefore, it is not applicable to the "Array" type.
Created attachment 199902 [details] Patch Introduce IsWrapperType subroutine.
Comment on attachment 199902 [details] Patch Looks good.
The commit-queue encountered the following flaky tests while processing attachment 199902 [details]: fast/frames/crash-remove-iframe-during-object-beforeload.html bug 115322 (author: zalan@apple.com) The commit-queue is continuing to process your patch.
Comment on attachment 199902 [details] Patch Clearing flags on attachment: 199902 Committed r149257: <http://trac.webkit.org/changeset/149257>
All reviewed patches have been landed. Closing bug.
Updated WebKitIDL wiki accordingly: http://trac.webkit.org/wiki/WebKitIDL?action=diff&version=90&old_version=89