350846 - Combine NS_***_LOAD events

Status: RESOLVED FIXED

(Core :: DOM: Events, defect)

Description

[Olli Pettay [:smaug][[e-mail]]]

I see no real reason to have 3 events for load.

Comment 1

[Olli Pettay [:smaug][[e-mail]]]

Attachment: proposed patch

I propose calling it just NS_LOAD.

(Removing all cases when multiple event types map to one event name helps with
the nsDOMEvent/nsEvent merge ... which I've been doing way too slowly.)

Comment 2

[Olli Pettay [:smaug][[e-mail]]]

Attachment: proposed patch.

Oops, the first patch was for some other bug.

Comment 3

[Johnny Stenback (:jst)]

Comment on attachment 236220 [details] [diff] [review]
proposed patch.

r+sr=jst, but I realized one thing as I was reviewing...

- In nsGlobalWindow::PostHandleEvent():

   } else if (aVisitor.mEvent->message == NS_PAGE_UNLOAD) {
     // Execute bindingdetached handlers before we tear ourselves
     // down.
     if (mDocument) {
       NS_ASSERTION(mDoc, "Must have doc");
       mDoc->BindingManager()->ExecuteDetachedHandlers();
     }
     mIsDocumentLoaded = PR_FALSE;
-  } else if (aVisitor.mEvent->message == NS_PAGE_LOAD) {
+  } else if (aVisitor.mEvent->message == NS_LOAD) {
+    // This is page load event since load events don't propagate to |window|.
+    // @see nsDocument::PreHandleEvent.
     mIsDocumentLoaded = PR_TRUE;

Both of these cases should also check if the event is trusted before setting the internal mIsDocumentLoaded state based on (un)load events. Unrelated to this patch, maybe, but I realized this as I was reviewing.

Comment 4

[Olli Pettay [:smaug][[e-mail]]]

Attachment: +NS_IS_TRUSTED_EVENT

I'll check in this.

Comment 5

[Boris Zbarsky [:bzbarsky]]

*** Bug 191961 has been marked as a duplicate of this bug. ***