Yeah, sounds good - it looks like you started splitting this into
separate changes already :)
I tried checking for metaKey in __updateControlKeys() and it didn't seem
to make any difference - but it's probably a good idea to check in both
places. We may need to refactor the stuck key detection code to its own
method...
André Bargull wrote:
I think we should/need to split up these changes to reduce testing
effort (because it's not only cross-browser, but also cross-OS).
So, I'd propose to change at first "LzKeyboardKernel.__keyboardEvent()"
and add "__updateControlKeys()". Which leads to my first question: Do
we need to check for 'metaKey' also "__updateControlKeys()" or just in
"__keyboardEvent()"?
On 6/2/2009 3:05 AM, Max Carlson wrote:
Change 20090601-maxcarlson-6 by maxcarl...@bank on 2009-06-01 17:44:36
PDT
in /Users/maxcarlson/openlaszlo/trunk-clean
for http://svn.openlaszlo.org/openlaszlo/trunk
Summary: Fix contextmenu inconsistencies in DHTML
Bugs Fixed: LPP-8218 - DHTML: issues with contextmenu onmenuopen,
dragging
Technical Reviewer: [email protected]
QA Reviewer: hminsky
Details: This includes all changes from
http://www.openlaszlo.org/pipermail/laszlo-dev/2009-May/021166.html -
Andre's changes to improve context menu support. I had to hand-merge
LzKeyboardKernel due to changes I'd made related to LPP-8210.
LzSprite - [Andre: canceling event-bubbling for all mouse-events =>
this was already done for all browsers (except Firefox) and all events
(except onmouseup). Why wasn't Firefox affected? => All other browsers
set "event.keyCode" to '0' for mouse-events, only Firefox sets
"keyCode" to 'undefined'. Now take a look at
"LzKeyboardKernel.__keyboardEvent()", every event was canceled in that
function if "keyCode" was '>= 0'!]
Moved detectstuckkeys from capabilities to quirks. Added
quirks.noglobaloncontextmenu to deal with Firefox OS X context menu
issue - https://bugzilla.mozilla.org/show_bug.cgi?id=443191.
__clickDispatcher() looks for quirks.noglobaloncontextmenu to fix
context menus in Firefox OS X. Remove unused (e.button == 2 &&
eventname != 'oncontextmenu') test from __mouseEvent().
LzKeyboardKernel - [Andre move handling for control keys (ctrl, shift,
alt) into separate function. It should be obvious why having a single
function for both, mouse-events and keyboard-events, is a bad idea
(see keyCode issue from above).] Hand-merged tests for meta key/stuck
keys checked into r14021.
LzMouseKernel - Andre's changes, plus I simplified the contextmenu
test to if (eventname == 'oncontextmenu' && targ) - testing for
(e.button == 2 && eventname == 'onmouseup') is now handled in
LzSprite.__clickDispatcher().
Tests: lz.Keys.downKeysArray no longer has extra entries for meta
keys, e.g. 'a' after pressing/releasing apple-a on OS X (LPP-8210).
test/contextmenu/lpp-6980.lzx?lzr=dhtml shows custom context menus for
red and blue boxes and default context menus for purple, yellow and
the canvas across Firefox and Safari OS X and Firefox, IE and Safari
on Windows. test/lfc/legals/keyboardandmouse.lzx?lzr=dhtml runs as
before in Firefox and
Safari OS X and Firefox, IE and Safari on Windows.
Files:
M WEB-INF/lps/lfc/kernel/dhtml/LzKeyboardKernel.js
M WEB-INF/lps/lfc/kernel/dhtml/LzSprite.js
M WEB-INF/lps/lfc/kernel/dhtml/LzMouseKernel.js
Changeset:
http://svn.openlaszlo.org/openlaszlo/patches/20090601-maxcarlson-6.tar
--
Regards,
Max Carlson
OpenLaszlo.org
