716207 - Cleanup in nsTextEditRules

Status: RESOLVED FIXED

(Core :: DOM: Editor, defect)

Description

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Attachment: Part a: DidUndo

Comment 1

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Attachment: Part b: RemoveRedundantTrailingBR

Comment 2

[(no longer active)]

Comment on attachment 586653 [details] [diff] [review]
Part b: RemoveRedundantTrailingBR

Review of attachment 586653 [details] [diff] [review]:
-----------------------------------------------------------------

::: editor/libeditor/text/nsTextEditRules.cpp
@@ +1077,4 @@
>  
> +  // Rather than deleting this node from the DOM tree we should instead
> +  // morph this br into the bogus node
> +  elem->UnsetAttr(kNameSpaceID_None, nsGkAtoms::type, /* ? */ true);

Why the doubt?  :-)

Comment 3

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Because the SetAttr right below it doesn't notify

Comment 4

[(no longer active)]

(In reply to Ms2ger from comment #3)
> Because the SetAttr right below it doesn't notify

Sure it does: <http://mxr.mozilla.org/mozilla-central/source/content/base/src/nsGenericElement.cpp#2575>

Comment 5

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Comment on attachment 586653 [details] [diff] [review]
Part b: RemoveRedundantTrailingBR

Yeah, the current code does notify, but I was wondering about this:

>+  // give it the bogus node attribute
>+  elem->SetAttr(kNameSpaceID_None, kMOZEditorBogusNodeAttrAtom,
>+                kMOZEditorBogusNodeValue, false);

Comment 6

[(no longer active)]

Yeah.  I don't know if we specifically _need_ to notify for removing the attribute (I think we don't) but I'd prefer this patch to not change that behavior.  I would be more than happy to review another patch in a new bug which turns off the notification.  That way we can deal with any possible future regressions better.  :-)

Comment 7

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

https://hg.mozilla.org/mozilla-central/rev/581a26455703
https://hg.mozilla.org/mozilla-central/rev/034af96fa579