53760 - paste pastes twice in Composer; undo undoes twice

Status: VERIFIED FIXED

(Core :: DOM: Editor, defect, P2)

Description

[Kathleen :Brade]

in both my Mac and Linux mozilla builds from today, pressing Command/Control-V to 
paste pastes the clipboard contents twice.  Undo does two steps of undo.  
Something is horribly broken.

Comment 1

[saari (gone)]

dr just fixed this

Comment 2

[Dan Rosen]

No, I didn't. That was different. I'm seeing this bug too, by the way...

This is probably a problem with having duplicate bindings for the same key.
Reassigning to saari to hack around in:

  xpfe/global/resources/content/htmlBindings.xml

and the platform html bindings. Either things that should only occur once are
occuring twice, or the new keylistener code is duplicating things that should
get overridden.

Comment 3

[Dan Rosen]

saari has a fix in hand. stupid things were happening. indicating smoketest
blocker. trying to get this in asap.

Comment 4

[Jean-Francois Ducarroz]

*** Bug 53775 has been marked as a duplicate of this bug. ***

Comment 5

[saari (gone)]

A diff for those that care

Index: nsXBLWindowKeyHandler.cpp
===================================================================
RCS file: /m/pub/mozilla/layout/xbl/src/nsXBLWindowKeyHandler.cpp,v
retrieving revision 1.2
diff -w -u -2 -r1.2 nsXBLWindowKeyHandler.cpp
--- nsXBLWindowKeyHandler.cpp	2000/09/22 05:31:36	1.2
+++ nsXBLWindowKeyHandler.cpp	2000/09/22 21:29:21
@@ -189,4 +189,12 @@
       return NS_OK;
  
+    nsCOMPtr<nsIDOMNSUIEvent> domUIEvent = do_QueryInterface(aKeyEvent);
+    if(domUIEvent){
+      PRBool prevent;
+      domUIEvent->GetPreventDefault(&prevent);
+      if (prevent)
+        return NS_OK;
+    }
+    
     PRBool matched = PR_FALSE;
     currHandler->KeyEventMatched(aEventType, aKeyEvent, &matched);
@@ -213,6 +221,8 @@
           rec = do_QueryInterface(elt);
         rv = currHandler->ExecuteHandler(rec, aKeyEvent);
-        if (NS_SUCCEEDED(rv))
+        if (NS_SUCCEEDED(rv)) {
+          aKeyEvent->PreventDefault();
           return NS_OK;
+        }
       }
     }

Comment 6

[Dan Rosen]

let's just say this isn't a smoketest blocker. i'd rather have people not worry
about this being on the radar during the branch. leaf instructed us to check in
as soon as the tree opens.

Comment 7

[Peter Trudelle]

nsbeta3+, P2 (due to workaround of using the menu)

Comment 8

[Jim Roskind]

Doing bad stuff to the document is much worse than doing nothing... so I'm
marking this as dogfood-plus.

Comment 9

[Jim Roskind]

Adding PDTP2 marking

Comment 10

[R.K.Aa.]

*** Bug 54002 has been marked as a duplicate of this bug. ***

Comment 11

[Dan Rosen]

PDT is spending a lot of time analysing this bug... I suppose I ought to have
made it clearer that we have a fix in hand :)

This will come as soon as we can check in.

Comment 12

[R.K.Aa.]

*** Bug 54056 has been marked as a duplicate of this bug. ***

Comment 13

[saari (gone)]

What PDT, or someone might want to do is ++ this bug...

Comment 14

[Peter Trudelle]

PDT agrees this is needed on branch, marking nsbeta3++.

Comment 15

[David Hyatt]

This diff is not quite right.  Let's talk about it on Tuesday.

Comment 16

[Peter Trudelle]

mass-adding rtm keyword to all open nsbeta3+ xptoolkit bugs

Comment 17

[R.K.Aa.]

*** Bug 54308 has been marked as a duplicate of this bug. ***

Comment 18

[rubydoo123]

*** Bug 54289 has been marked as a duplicate of this bug. ***

Comment 19

[saari (gone)]

here is the hyatt approved diff

Index: nsXBLWindowKeyHandler.cpp
===================================================================
RCS file: /m/pub/mozilla/layout/xbl/src/nsXBLWindowKeyHandler.cpp,v
retrieving revision 1.2
diff -w -u -2 -r1.2 nsXBLWindowKeyHandler.cpp
--- nsXBLWindowKeyHandler.cpp	2000/09/22 05:31:36	1.2
+++ nsXBLWindowKeyHandler.cpp	2000/09/28 01:01:26
@@ -213,6 +213,8 @@
           rec = do_QueryInterface(elt);
         rv = currHandler->ExecuteHandler(rec, aKeyEvent);
-        if (NS_SUCCEEDED(rv))
+        if (NS_SUCCEEDED(rv)) {
+          aKeyEvent->PreventDefault();
           return NS_OK;
+        }
       }
     }

Comment 20

[saari (gone)]

Fix in for both branch and trunk

Comment 21

[sujay]

verified in 9/28 build.