SliderThumbElement and SpinButtonElement's isEnabledFormControl are the only ones that do not just return !disabled(). If we can make them return !disabled(), it would prove that having both disabled and isEnabledFormControl is redundant, and we can just refactor them into a single function.
Created attachment 194420 [details] Patch
Comment on attachment 194420 [details] Patch ok
Comment on attachment 194420 [details] Patch Clearing flags on attachment: 194420 Committed r146638: <http://trac.webkit.org/changeset/146638>
All reviewed patches have been landed. Closing bug.
Isn’t disabled is a web-exposed property? If so, then it seems that this change is a web-exposed change in behavior, not just a refactoring. If that is right, then we need a regression test and we need to be sure this is a correct change.
(In reply to comment #5) > Isn’t disabled is a web-exposed property? If so, then it seems that this change is a web-exposed change in behavior, not just a refactoring. If that is right, then we need a regression test and we need to be sure this is a correct change. No. Web-exposed 'disabled' IDL attributes are [Reflected]. Also, SliderThumbElement and SpinButtonElement are not Web-exposed.
(In reply to comment #6) > Web-exposed 'disabled' IDL attributes are [Reflected]. Aha! Then it is critical to eliminate or rename the C++ member function named disabled since it is not the same as the web-exposed function named disabled, which is super-confusing!
(In reply to comment #7) > Then it is critical to eliminate or rename the C++ member function named disabled since it is not the same as the web-exposed function named disabled, which is super-confusing! Indeed. I'd like to rename HTMLFormControlElement::readOnly too because of the same reason.
(In reply to comment #7) > Then it is critical to eliminate or rename the C++ member function named disabled since it is not the same as the web-exposed function named disabled, which is super-confusing! I'm hoping to do this: bug 112085, comment 13