Add a warning when tuplet.start_note.voice != tuplet.end_note.voice

[leleogere]

As discussed in #420, the start and end notes of a tuplet are a good indicator of an XML encoding error. This happens when there is a missing tag <tuplet type="start"> or <tuplet type="stop">. For instance, it is the case in the score Chopin/Sonata_2/1st_no_repeat in ASAP (see fix at fosfrancesco/asap-dataset#14).

This PR adds a warning when this happens, providing information to the user that there might be a missing tuplet tag.

/home/user/partitura/partitura/io/importmusicxml.py:1577: UserWarning: Tuplet start and end notes do not belong to the same voice (5 != 6). This might indicate a missing <tuplet type="start"> or <tuplet type="stop">.
  warnings.warn(

MuseScore is able to fix this kind of encoding error, should we try too? Here is an overview of a naive approach presented in #420:

• Find an additional <tuplet type="stop"> without pending <tuplet type="start"> registered.
• In that case, iterate backward over previous notes of the same voice, as long as they have the same <time-> modification> as the "end" note, and they do not have any <tuplet> attribute.
  • Should we search only on the current measure? Tuplets crossing bar lines are extremely rare;
• Set the start_note of the tuplet to the last note verifying the previous conditions.

In case of missing <tuplet type="stop">, the same could be applied but iterating forward from the <tuplet > type="start"> rather than backward. However, detecting missing stops might be a bit more difficult. Indeed, in the case of a missing start, you detect it right away when you encounter a stop without having found a start before, but for missing stops, you need some rules that say "the tuplet should be ended by now, there must be a stop missing".

However, there might be more complex cases that wouldn't be covered by this approach (I think of nested tuplets, but there might be other weird cases).

---

For reference, here is a related part of the code of MuseScore:
https://github.com/musescore/MuseScore/blob/559bf1eb46e467a000a64a36b13173f3e58e2a32/src/importexport/musicxml/internal/musicxml/import/musicxmltupletstate.cpp#L344-L366

    // Tuplets are stopped by the tuplet stop
    // or when the tuplet is filled completely
    // (either with knowledge of the normal type
    // or as a last resort calculated based on
    // actual and normal notes plus total duration)
    // or when the time-modification is not found.

    if (inTuplet) {
        if (tupletStartStop == MusicXmlStartStop::STOP
            || (implicit && isTupletFilled(normalType, timeMod))
            || (actNotes == 1 && norNotes == 1)) {           // incorrect ??? check scenario incomplete tuplet w/o start
            if (actNotes > norNotes && !isTupletFilled(normalType, timeMod)) {
                missingCurrentDuration = missingTupletDuration(duration);
                LOGD("current tuplet incomplete, missing %s", muPrintable(missingCurrentDuration.toString()));
            }

            *this = {};
            res |= MusicXmlTupletFlag::STOP_CURRENT;
        }
    }

    return res;
}

I'm not really used to C++, but it might be worth to have a look at what's happening in here.
One thing they seem to do (among others) it to keep track of the time in the tuplet, which might be a lot of work to include in Partitura if we choose that way.

[manoskary]

This PR addresses potential errors in tuplets by raising a warning, it does not propose a solution or re-assignment but rather notifies the user about potentially problematic scores.