MusicXML import with 2 voices per staff get confused by crescendo spanners

[Tobias Kretschmar]

> I'm not top posting.

Given is a MusicXML file with 2 voices on one staff. Voice one contains 
a crescendo spanner.

musicxml2ly.py will create a ly file with faulty skips (hidden rests) or 
notes scrambled in chords.

The reason is in musicxml.py line 1309 a direction wedge is added to any 
voice. This will lead to following voice elements:

Voice1:
1. Note1 half at when=0
2. Note2 half at when=1/2
3. Direction wedge crescendo at when=1/2


Voice2:
1. Direction wedge crescendo at when=1/2
2. Note1 quarter at when=0
3. Note2 quarter at when=1/2


Subsequently in musicxml2ly line 2122 voice_bilder.jumpto (n.when) is 
called for any event in Voice1 which is ok and then in Voice2.

1. Direction wedge at when 1/2: will insert a skip 1/2
2. Note1 at when 0 will throw a warning about negative skip -1/2
3. Note2 at 1/2, ok

Solution: I assume, direction wedges have to be coupled to specific 
voices?

Example musicxml file:
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE score-partwise PUBLIC "-//Recordare//DTD MusicXML 3.0 
Partwise//EN" "http://www.musicxml.org/dtds/partwise.dtd";>
<score-partwise>
  <part-list>
    <score-part id="P1">
      </score-part>
    </part-list>
  <part id="P1">
    <measure number="1" width="146.59">
      <attributes>
        <divisions>2</divisions>
        <key>
          <fifths>-1</fifths>
        </key>
        <time symbol="cut">
          <beats>2</beats>
          <beat-type>2</beat-type>
        </time>
        <clef>
          <sign>G</sign>
          <line>2</line>
        </clef>
      </attributes>
      <note default-x="12.00" default-y="-25.00">
        <pitch>
          <step>A</step>
          <octave>4</octave>
          </pitch>
        <duration>4</duration>
        <voice>1</voice>
        <type>half</type>
        <stem>up</stem>
        </note>
      <direction placement="below">
        <direction-type>
          <wedge type="crescendo" number="1"/>
          </direction-type>
        </direction>
      <note default-x="70.94" default-y="-25.00">
        <pitch>
          <step>B</step>
          <octave>4</octave>
          </pitch>
        <duration>4</duration>
        <voice>1</voice>
        <type>half</type>
        <stem>up</stem>
        </note>
      <direction placement="below">
        <direction-type>
          <wedge type="stop" number="1"/>
        </direction-type>
      </direction>
      <backup>
        <duration>8</duration>
        </backup>
      <note default-x="12.00" default-y="-35.00">
        <pitch>
          <step>F</step>
          <octave>4</octave>
          </pitch>
        <duration>4</duration>
        <voice>2</voice>
        <type>half</type>
        <stem>down</stem>
        </note>
      <note default-x="70.94" default-y="-35.00">
        <pitch>
          <step>G</step>
          <octave>4</octave>
          </pitch>
        <duration>4</duration>
        <voice>2</voice>
        <type>half</type>
        <stem>down</stem>
        </note>
      </measure>
    </part>
  </score-partwise>

Output ly file:

\version "2.19.16"
% automatically converted by musicxml2ly from 
C:/Users/Tobi/Documents/Visual Studio 
2013/Projects/PythonApplication1/PythonApplication1/test.xml

%% additional definitions required by the score:
\language "deutsch"


\header {
}

PartPOneVoiceOne =  \relative a' {
  \transposition b \clef "treble" \key f \major \time 2/2 a2 h \< \!
}

PartPOneVoiceTwo =  \relative f' {
  \transposition b \clef "treble" \key f \major \time 2/2 s2 s \< f
  \! g2
}


% The score definition
\score {
  <<
    \new Staff <<
      \context Staff <<
        \context Voice = "PartPOneVoiceOne" { \voiceOne 
\PartPOneVoiceOne }
        \context Voice = "PartPOneVoiceTwo" { \voiceTwo 
\PartPOneVoiceTwo }
      >>
    >>

  >>
  \layout {}
  \midi {}
}


Expected output:

\version "2.19.16"
% automatically converted by musicxml2ly from 
C:/Users/Tobi/Documents/Visual Studio 
2013/Projects/PythonApplication1/PythonApplication1/test.xml

%% additional definitions required by the score:
\language "deutsch"


\header {
}

PartPOneVoiceOne =  \relative a' {
  \transposition b \clef "treble" \key f \major \time 2/2 a2 h \< \!
}

PartPOneVoiceTwo =  \relative f' {
  \transposition b \clef "treble" \key f \major \time 2/2 f2 g
}


% The score definition
\score {
  <<
    \new Staff <<
      \context Staff <<
        \context Voice = "PartPOneVoiceOne" { \voiceOne 
\PartPOneVoiceOne }
        \context Voice = "PartPOneVoiceTwo" { \voiceTwo 
\PartPOneVoiceTwo }
      >>
    >>

  >>
  \layout {}
  \midi {}
}