avoid dots for notes which may result in 256th notes/rests or shorter

[AntonioBL]

This is just a tentative implementation to avoid problems such as:
http://musescore.org/node/23473
http://musescore.org/en/node/9077
I don't know if there may be other possible extreme cases in which these short-notes-problems may arise.

[lasconic]

What about disabling the dot and dotdot buttons (and actions) when a short note is selected instead?

[AntonioBL]

Well... That's what I first wanted to do, but I couldn't find the right place where to disable the button.
After this PR, I tried a couple of experiments, but the results are not 100% satisfactory. The one that seems to work best is to replace lines 308 e 309 of mscore\keyb.cpp

      getAction("pad-dot")->setChecked(is.duration().dots() == 1);
      getAction("pad-dotdot")->setChecked(is.duration().dots() == 2);

with:

      if (is.duration() == TDuration::V_128TH) {
            getAction("pad-dot")->setChecked(false);
            getAction("pad-dot")->setEnabled(false);
            getAction("pad-dotdot")->setChecked(false);
            getAction("pad-dotdot")->setEnabled(false);
            }
      else if (is.duration() == TDuration::V_64TH) {
            getAction("pad-dotdot")->setChecked(false);
            getAction("pad-dotdot")->setEnabled(false);
            getAction("pad-dot")->setEnabled(true);
            getAction("pad-dot")->setChecked(is.duration().dots() == 1);
            }
      else {
            getAction("pad-dot")->setEnabled(true);
            getAction("pad-dotdot")->setEnabled(true);
            getAction("pad-dot")->setChecked(is.duration().dots() == 1);
            getAction("pad-dotdot")->setChecked(is.duration().dots() == 2);
            }

That is not a really elegant implementation, maybe the code can be written better.
Moreover, when a 128th (or 64th) note/rest is already present in the score, if it is selected before entering note-entry mode, the buttons become enabled (but not checked): clicking over them disables them [without checking them and therefore without creating a dotted 128th: that's the purpose of the setChecked(false) ].
I didn't check too much these lines of code: I don't know if there may be collateral strange effects.
Do you think it's worth a pull request for this code?

[AntonioBL]

Just to precede @Jojo-Schmitz in asking: should be PR be closed? :-)
Did you try the alternative implementation proposed in the comment? Do you think it would be worth a PR by its own?
Ciao,
ABL

[Jojo-Schmitz]

:-)

[AntonioBL]

Updated and rebased the PR. It now seems to work as expected (i.e. it disables double dot for 1/64 and 1/128 notes and dot for 1/128 notes).

[lasconic]

I merged but it doesn't work I expected. See http://musescore.org/en/node/50386
I'll revert.