Added ripple delete #6692
Added ripple delete #6692
Conversation
|
When the selection of notes is not "continuous" (Shift+Ctrl+Mouse) the behavior is not consistent. Not sure what the behavior should be. Perhaps treat each "continuous" sub-group individually. |
|
@superpaik I was not sure about this behavior either when developing it. I assumed that if a note (or notes) was within the range of the selected notes but not actually selected, as what it sounds like you are alluding to above, the unselected note(s) should not be shifted upon ripple deleting the selected notes. So, sort of ignoring those notes was an intentional choice since this behavior was not specified in the feature request. I think your suggestion on what to do in this case would make sense, however implementing it may be a bit tricky in trying to detect notes in between the gaps of selected notes. I think I would have to partition the selected notes into different regions, detect the unselected notes between them, and then shift the unselected notes between them based on the length of the partitions. It seems feasible, but again, a bit tricky, especially since I would probably have to restructure some of the code I already wrote. Do you want to keep the behavior as it currently is or do you want me to try and make this adjustment? |
|
@hberg137 As I said, not so sure what the best behavior should be. I recognize it's a weird use-case scenario. If it was intended and as you had a discussion already, maybe the current is the best possible behavior. |
No please. This creates very weird results. I see two feasible alternatives...
|
| deleteSelectedNotes(); | ||
| ke->accept(); | ||
| break; | ||
| { |
There was a problem hiding this comment.
Could we move all this code into deleteSelectedNotes() instead, and just pass it something like move=true ?
| if (selectedNotes.empty()) { break; } | ||
| TimePos delStartPos = selectedNotes.first()->pos(); | ||
| TimePos delEndPos = selectedNotes.last()->endPos(); | ||
| TimePos delEndLen = selectedNotes.last()->length(); |
There was a problem hiding this comment.
A prior note may still end after the last note depending on its length. You may find boundsForNotes helpful.
It will still be tricky with drum notes, but I imagine something like this:
area to delete |xxxxxx| |xxxxxxxxx|
normal note ###### #### #### ###
drum note # #
In that case you just need to check if the drum note starts on the end position... I also see that boundsForNotes doesn't handle drum notes well. That needs to be fixed somehow.
| delEndPos = rightNotes.first()->pos(); | ||
| } | ||
|
|
||
| shiftPos(rightNotes, delStartPos - delEndPos); |
There was a problem hiding this comment.
You may disable journalling before moving to prevent the double undo
m_midiClip->saveJournallingState(false);
shiftPos(...);
m_midiClip->restoreJournallingState();
|
Any progress on this ? |
Pull request for ripple delete issue #6495
Please see the following video for demonstration/explanation.