mart added a comment.

  In https://phabricator.kde.org/D8641#167459, @hein wrote:
  
  > So if I understand correctly, the Mnemonic stuff currently doesn't seem to 
have the same aims as KAcceleratorManager does ...
  >
  > KAcceleratorManager acts on a hierarchy of widgets, traverses it, assigns 
weights to actionable things and possible mnemonics and overall smartly tries 
to distribute mnemonics across the hierarchy.
  >
  > If I understand correctly this is more like a convenience wrapper around 
QKeySequence::mnemonic() that you can get both the label and a sequence to pass 
to shortcut to.
  
  
  no, it's kindof like kacceleratormanager.. but much more primitive, as it has 
just a single global pool in which accelerators are unique.
  
  so what it would need and i may try, is to make and hash per different 
window, and then  have something about widgets hyerarchies.
  tough i was thinking that may be enough if is global per-window and shortcuts 
are unique based on only one enabled at the moment (so like if there is a 
stackview, only the topmost current page would be enabled)
  
  > I'm not really sure about those static mappings members ... there currently 
doesn't seem to be any duplicate mnemonic tracking yet, right?
  
  basically it just avoids any duplicates... first come first served
  
  > My feeling is that this is still too much boiler plate for the dev. At 
least within a FormLayout, which has knowledge of what labels go with what, all 
the assignment should probably just happen automatically without needing to 
have your own Shortcut {} instances and getting the labels out of the Mnemonic 
thing. Do you think we can do that somehow?
  
  there would be boiler plate for who writes their own controls (which oone 
should try hard to avoid anyways)
  using a formlayout just works and for primitive controls should just work 
automatically as well, if this is done in the style.
  
  > Overall the code in KAcceleratorManagement is quite good and nicely 
commented, there's likely stuff of value in there (e.g. the algorithm it uses 
to weight mnemonic characters).
  
  i'll look at the algorythm to see if is reusable

REPOSITORY
  R169 Kirigami

REVISION DETAIL
  https://phabricator.kde.org/D8641

To: mart, #plasma, #kirigami, hein
Cc: colomar, ngraham, davidedmundson, plasma-devel, ZrenBot, progwolff, 
lesliezhai, ali-mohamed, jensreuterberg, abetts, sebas, apol, mart, hein

Reply via email to