Bug 37277 – Move the IDB::open ExceptionCode parameter to be last

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Jeremy Orlow]

Move the IDB::open ExceptionCode paramter to be last

[Jeremy Orlow]

Created attachment 52873 [details]
Patch

[Darin Adler]

Normally we use the typedef ExceptionCode so we don't have to name exception code arguments. Please do that whenever practical.

[Jeremy Orlow]

(In reply to comment #2)
> Normally we use the typedef ExceptionCode so we don't have to name exception
> code arguments. Please do that whenever practical.

All the uses of int are in the Chromium WebKit API.  I suppose maybe we should add a WebExceptionCode class there, but I also don't see a huge advantage to doing so.

+ fishd in case we think it's a good idea to do so.

[Darin Adler]

(In reply to comment #3)
> All the uses of int are in the Chromium WebKit API.  I suppose maybe we should
> add a WebExceptionCode class there, but I also don't see a huge advantage to
> doing so.

I'm surprised that ExceptionCode is part of the Chromium WebKit API. It's an implementation detail we've often changed around in the past and not something particularly stable, so not great for API. But anyway, not on point for this patch.

[Jeremy Orlow]

(In reply to comment #4)
> (In reply to comment #3)
> > All the uses of int are in the Chromium WebKit API.  I suppose maybe we should
> > add a WebExceptionCode class there, but I also don't see a huge advantage to
> > doing so.
> 
> I'm surprised that ExceptionCode is part of the Chromium WebKit API. It's an
> implementation detail we've often changed around in the past and not something
> particularly stable, so not great for API. But anyway, not on point for this
> patch.

Did you mean you're surprised it's not?

Anyway, I guess you're right that it is an implementation detail that should probably be abstracted away by an API.  I'll look into changing this, unless you have an objection, fishd.

[Darin Adler]

(In reply to comment #5)
> > I'm surprised that ExceptionCode is part of the Chromium WebKit API. It's an
> > implementation detail we've often changed around in the past and not something
> > particularly stable, so not great for API. But anyway, not on point for this
> > patch.
> 
> Did you mean you're surprised it's not?

I am surprised that WebCore's internal representation of DOM exceptions is a part of the API. Not the ExceptionCode typedef, which, as you say, is not.

[WebKit Commit Bot]

Comment on attachment 52873 [details]
Patch

Clearing flags on attachment: 52873

Committed r57350: <http://trac.webkit.org/changeset/57350>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.