Source code for music21.abcFormat.translate

# -----------------------------------------------------------------------------
# Name:         abcFormat.translate.py
# Purpose:      Translate ABC and music21 objects
#
# Authors:      Christopher Ariza
#               Michael Scott Asato Cuthbert
#               Dylan Nagler
#
# Copyright:    Copyright © 2010-2024 Michael Scott Asato Cuthbert
# License:      BSD, see license.txt
# -----------------------------------------------------------------------------
'''
Functions for translating music21 objects and
:class:`~music21.abcFormat.ABCHandler` instances.
Mostly, these functions are for advanced, low level usage.
For basic importing of ABC files from a file or URL to a
:class:`~music21.stream.Stream`, use the music21 converter
module's :func:`~music21.converter.parse` function.
'''
from __future__ import annotations

import copy
import unittest
import re
import typing  # cannot import as t because of use of t as token throughout

from music21 import articulations
from music21 import bar
from music21 import clef
from music21 import chord
from music21 import common
from music21 import environment
from music21 import exceptions21
from music21 import harmony
from music21 import meter
from music21 import note
from music21 import spanner
from music21 import stream
from music21 import tempo
from music21 import tie

if typing.TYPE_CHECKING:
    from music21 import abcFormat


environLocal = environment.Environment('abcFormat.translate')

_abcArticulationsToM21 = {
    'staccato': articulations.Staccato,
    'upbow': articulations.UpBow,
    'downbow': articulations.DownBow,
    'accent': articulations.Accent,
    'strongaccent': articulations.StrongAccent,
    'tenuto': articulations.Tenuto,
}


[docs] def abcToStreamPart( abcHandler: abcFormat.ABCHandler, inputM21: stream.Part|None = None, spannerBundle: spanner.SpannerBundle|None = None ) -> stream.Part: ''' Handler conversion of a single Part of a Score with multiple Parts. Results are added into the provided inputM21 object or a newly created Part object. The part object is then returned. ''' from music21 import abcFormat if inputM21 is None: p = stream.Part() else: p = inputM21 if spannerBundle is None: # environLocal.printDebug(['mxToMeasure()', 'creating SpannerBundle']) spannerBundle = spanner.SpannerBundle() clefSet = None postTransposition = 0 dst: stream.Part|stream.Measure mergedHandlers: list[abcFormat.ABCHandler] = [] # need to call on entire handlers, as looks for special criteria, # like that at least 2 regular bars are used, not just double bars if abcHandler.definesMeasures(): # first, split into a list of Measures; if there is only metadata and # one measure, that means that no measures are defined barHandlers = abcHandler.splitByMeasure() # environLocal.printDebug(['barHandlers', len(barHandlers)]) # merge loading metadata with each bar that precedes it mergedHandlers = typing.cast(list[abcFormat.ABCHandler], abcFormat.mergeLeadingMetaData(barHandlers)) # environLocal.printDebug(['mergedHandlers', len(mergedHandlers)]) else: # simply stick in a single list mergedHandlers = [abcHandler] # if only one merged handler, do not create measures if len(mergedHandlers) <= 1: useMeasures = False else: useMeasures = True # each unit in merged handlers defines possible a Measure (w/ or w/o metadata), # trailing metadata, or a single collection of metadata and note data barCount = 0 measureNumber = 1 # merged handler are ABCHandlerBar objects, defining attributes for barlines for mh in mergedHandlers: if typing.TYPE_CHECKING: assert isinstance(mh, abcFormat.ABCHandlerBar) # if "use measures" is True and the handler has notes; otherwise add to part # environLocal.printDebug(['abcToStreamPart', 'handler', 'left:', mh.leftBarToken, # 'right:', mh.rightBarToken, 'len(mh)', len(mh)]) if useMeasures and mh.hasNotes(): # environLocal.printDebug(['abcToStreamPart', 'useMeasures', # useMeasures, 'mh.hasNotes()', mh.hasNotes()]) dst = stream.Measure() # bar tokens are already extracted from token list and are available # as attributes on the handler object # may return None for a regular barline if mh.leftBarToken is not None: # this may be Repeat Bar subclass bLeft = mh.leftBarToken.getBarObject() if bLeft is not None: dst.leftBarline = bLeft if mh.leftBarToken.isRepeatBracket(): # get any open spanners of RepeatBracket type rbSpanners = spannerBundle.getByClass(spanner.RepeatBracket ).getByCompleteStatus(False) # this indication is most likely an opening, as ABC does # not encode second ending boundaries # we can still check thought: if not rbSpanners: # add this measure as a component rb = spanner.RepeatBracket(dst) # set number, returned here rb.number = mh.leftBarToken.isRepeatBracket() # only append if created; otherwise, already stored spannerBundle.append(rb) else: # close it here rb = typing.cast(spanner.RepeatBracket, rbSpanners[0]) # get RepeatBracket rb.addSpannedElements(dst) rb.completeStatus = True # this returns 1 or 2 depending on the repeat # in ABC, second repeats close immediately; that is # they never span more than one measure if mh.leftBarToken.isRepeatBracket() == 2: rb.completeStatus = True if mh.rightBarToken is not None: bRight = mh.rightBarToken.getBarObject() if bRight is not None: dst.rightBarline = bRight # above returns bars and repeats; we need to look if we just # have repeats if mh.rightBarToken.isRepeat(): # if we have a right bar repeat, and a spanner repeat # bracket is open (even if just assigned above) we need # to close it now. # presently, now r bar conditions start a repeat bracket rbSpanners = spannerBundle.getByClass( spanner.RepeatBracket).getByCompleteStatus(False) if any(rbSpanners): rb = typing.cast(spanner.RepeatBracket, rbSpanners[0]) rb.addSpannedElements(dst) rb.completeStatus = True # this returns 1 or 2 depending on the repeat # do not need to append; already in bundle barCount += 1 else: dst = p # store directly in a part instance # environLocal.printDebug([mh, 'dst', dst]) # ql = 0 # might not be zero if there is a pickup postTransposition, clefSet = parseTokens(mh, dst, p, useMeasures) # append measure to part; in the case of trailing metadata # dst may be part, even though useMeasures is True if useMeasures and isinstance(dst, stream.Measure): # check for incomplete bars # must have a time signature in this bar, or defined recently # could use getTimeSignatures() on Stream if barCount == 1 and dst.timeSignature is not None: # easy case # can only do this b/c ts is defined if dst.barDurationProportion() < 1.0: dst.padAsAnacrusis() dst.number = 0 # environLocal.printDebug([ # 'incompletely filled Measure found on abc import; ', # 'interpreting as a anacrusis:', 'paddingLeft:', dst.paddingLeft]) else: dst.number = measureNumber measureNumber += 1 p.coreAppend(dst) p.coreElementsChanged() try: reBar(p, inPlace=True) except (ABCTranslateException, meter.MeterException, ZeroDivisionError): pass # clefs are not typically defined, but if so, are set to the first measure # following the metadata, or in the open stream if not clefSet and not p[clef.Clef]: if useMeasures: # assume at start of measures m = p.getElementsByClass(stream.Measure).first() if typing.TYPE_CHECKING: assert m is not None m.clef = clef.bestClef(p, recurse=True) else: p.insert(0, clef.bestClef(p, recurse=True)) if postTransposition != 0: p.transpose(postTransposition, inPlace=True) if useMeasures and p[meter.TimeSignature]: # call make beams for now; later, import beams # environLocal.printDebug(['abcToStreamPart: calling makeBeams']) try: p.makeBeams(inPlace=True) except (meter.MeterException, stream.StreamException) as e: environLocal.warn(f'Error in beaming. Ignoring: {e}') # copy spanners into topmost container; here, a part rm = [] for sp in spannerBundle.getByCompleteStatus(True): p.coreInsert(0, sp) rm.append(sp) # remove from original spanner bundle for sp in rm: spannerBundle.remove(sp) p.coreElementsChanged() # first_ts = p[meter.TimeSignature].first() # ts_ql = first_ts.barDuration.quarterLength if first_ts else 4.0 # if p.highestTime >= ts_ql * 3: # p = p.makeMeasures() return p
[docs] def parseTokens( mh: abcFormat.ABCHandler, dst: stream.Measure|stream.Part, p: stream.Part, useMeasures: bool ) -> tuple[int, bool]: ''' Parses all the tokens in a measure or part. ''' # in case need to transpose due to clef indication from music21 import abcFormat postTransposition = 0 clefSet = False for t in mh.tokens: if isinstance(t, abcFormat.ABCMetadata): postTransposition, clefSet = metadataToM21Object( t, dst, postTransposition, clefSet, useMeasures ) elif isinstance(t, abcFormat.ABCNote): parseABCNote(t, dst) elif isinstance(t, abcFormat.ABCSlurStart): if typing.TYPE_CHECKING: assert t.slurObj is not None p.coreAppend(t.slurObj) elif isinstance(t, abcFormat.ABCCrescStart): if typing.TYPE_CHECKING: assert t.crescObj is not None p.coreAppend(t.crescObj) elif isinstance(t, abcFormat.ABCDimStart): if typing.TYPE_CHECKING: assert t.dimObj is not None p.coreAppend(t.dimObj) dst.coreElementsChanged() return postTransposition, clefSet
[docs] def parseABCNote( t: abcFormat.ABCNote, dst: stream.Measure|stream.Part, ) -> None: ''' Parse an ABCNote object and add it to the destination stream. ''' from music21 import abcFormat n: note.GeneralNote cs: harmony.ChordSymbol|harmony.NoChord # add the attached chord symbol if t.chordSymbols: cs_name = t.chordSymbols[0] cs_name = re.sub('"', '', cs_name).lstrip().rstrip() cs_name = re.sub('[()]', '', cs_name) cs_name = common.cleanedFlatNotation(cs_name) try: if cs_name in ('NC', 'N.C.', 'No Chord', 'None'): cs = harmony.NoChord(cs_name) elif cs_name.startswith('>'): return # fingering diagram? Appears in some pieces, ryans-Neumedia else: cs = harmony.ChordSymbol(cs_name) dst.coreAppend(cs, setActiveSite=False) dst.coreElementsChanged() except ValueError: pass # Exclude malformed chord # as ABCChord is subclass of ABCNote, handle first if isinstance(t, abcFormat.ABCChord): # Skip an empty chord if not t.subTokens: return # may have more than notes? pitchNameList: list[str] = [] accStatusList = [] # accidental display status list for tSub in t.subTokens: # notes are contained as subTokens are already parsed if isinstance(tSub, abcFormat.ABCNote): pn = tSub.pitchName if typing.TYPE_CHECKING: assert pn is not None pitchNameList.append(pn) accStatusList.append(tSub.accidentalDisplayStatus) c = chord.Chord(pitchNameList) c.duration.quarterLength = t.quarterLength if t.activeTuplet: thisTuplet = copy.deepcopy(t.activeTuplet) if thisTuplet.durationNormal is None: thisTuplet.setDurationType(c.duration.type, c.duration.dots) c.duration.appendTuplet(thisTuplet) # adjust accidental display for each contained pitch for pIndex in range(len(c.pitches)): acc = c.pitches[pIndex].accidental if acc is not None: acc.displayStatus = accStatusList[pIndex] dst.coreAppend(c) # ql += t.quarterLength else: if t.isRest: n = note.Rest() else: n = note.Note(t.pitchName) if n.pitch.accidental is not None: n.pitch.accidental.displayStatus = t.accidentalDisplayStatus n.duration.quarterLength = t.quarterLength if t.activeTuplet: thisTuplet = copy.deepcopy(t.activeTuplet) if thisTuplet.durationNormal is None: thisTuplet.setDurationType(n.duration.type, n.duration.dots) n.duration.appendTuplet(thisTuplet) # start or end a tie at note n if t.tie is not None: if t.tie in ('start', 'continue'): n.tie = tie.Tie(t.tie) n.tie.style = 'normal' elif t.tie == 'stop': n.tie = tie.Tie(t.tie) # Was: Extremely Slow for large Opus files. Why? # Answer: some pieces didn't close all their spanners, so # everything was in a Slur/Diminuendo, etc. for span in t.applicableSpanners: span.addSpannedElements(n) if t.inGrace: n = n.getGrace() n.articulations = [] while any(t.articulations): tokenArticulationStr = t.articulations.pop() if tokenArticulationStr not in _abcArticulationsToM21: continue m21ArticulationClass = _abcArticulationsToM21[tokenArticulationStr] m21ArticulationObj = m21ArticulationClass() n.articulations.append(m21ArticulationObj) dst.coreAppend(n, setActiveSite=False)
[docs] def metadataToM21Object( t: abcFormat.ABCMetadata, dst: stream.Measure|stream.Part, postTransposition: int, clefSet: bool, useMeasures: bool, ) -> tuple[int, bool]: ''' Parse a single ABCMetadata token, that does not represent what music21 considers metadata, such as a time signature, key signature, clef, or tempo, and put it in dst. This function uses coreAppend operations, so if called separately, dst.coreElementsChanged() should be called before accessing the stream. Returns the changed transposition and whether clef was set. ''' if t.isMeter(): ts = t.getTimeSignatureObject() if ts is not None: # can be None # should append at the right position if useMeasures: # assume at start of measures dst.timeSignature = ts else: dst.coreAppend(ts) elif t.isKey(): ks = t.getKeySignatureObject() if ks is not None and useMeasures: # assume at start of measures dst.keySignature = ks elif ks is not None: dst.coreAppend(ks) # check for clef information sometimes stored in key clefObj, transposition = t.getClefObject() if clefObj is not None and transposition is not None: # "and transposition is not None" for type checking. clefSet = False # environLocal.printDebug(['found clef in key token:', t, # clefObj, transposition]) if useMeasures: # assume at start of measures dst.clef = clefObj else: dst.coreAppend(clefObj) postTransposition = transposition elif t.isTempo(): mmObj = t.getMetronomeMarkObject() if typing.TYPE_CHECKING: assert mmObj is not None dst.coreAppend(mmObj) return postTransposition, clefSet
[docs] def abcToStreamScore( abcHandler: abcFormat.ABCHandler, inputM21: stream.Score|None = None ) -> stream.Score: ''' Given an abcHandler object, build into a multi-part :class:`~music21.stream.Score` with metadata. This assumes that this ABCHandler defines a single work (with 1 or fewer reference numbers). if the optional parameter inputM21 is given a music21 Stream subclass, it will use that object as the outermost object. However, inner parts will always be made :class:`~music21.stream.Part` objects. ''' from music21 import abcFormat from music21 import metadata if inputM21 is None: s = stream.Score() else: s = inputM21 # metadata can be first md = metadata.Metadata() s.insert(0, md) # get title from large-scale metadata titleCount = 0 for t in abcHandler.tokens: if isinstance(t, abcFormat.ABCMetadata): if t.isVersion(): v = t.data.replace('abc-version', '').strip() versionMatch = re.match(r'(\d+).(\d+).?(\d+)?', v) if versionMatch is not None: abcHandler.abcVersion = abcHandler.returnAbcVersionFromMatch( versionMatch ) elif t.isTitle(): if titleCount == 0: # first md.add('title', t.data) # environLocal.printDebug(['got metadata title', t.data]) titleCount += 1 # all other titles go in alternative field else: md.add('alternativeTitle', t.data) # environLocal.printDebug(['got alternative title', t.data]) titleCount += 1 elif t.isComposer(): md.add('composer', t.data) # environLocal.printDebug(['got composer', t.data]) elif t.isOrigin(): md.add('localeOfComposition', t.data) # environLocal.printDebug(['got locale of composition', t.data]) elif t.isReferenceNumber(): md.add('number', int(t.data)) # environLocal.printDebug(['got work number', t.data]) partHandlers = [] tokenCollections = abcHandler.splitByVoice() if len(tokenCollections) == 1: partHandlers.append(tokenCollections[0]) else: # add metadata -- stored in tokenCollections[0] -- # to each Part (stored in tokenCollections[i]) for i in range(1, len(tokenCollections)): # concatenate abc handler instances newABCHandler = tokenCollections[0] + tokenCollections[i] # dummy = [t.src for t in newABCHandler.tokens] # print(dummy) partHandlers.append(newABCHandler) # find if this token list defines measures # this should probably operate at the level of tunes, not the entire # token list partList = [] for partHandler in partHandlers: p = abcToStreamPart(partHandler) partList.append(p) for p in partList: s.coreInsert(0, p) s.coreElementsChanged() return s
[docs] def abcToStreamOpus( abcHandler: abcFormat.ABCHandler, inputM21: stream.Opus|None = None, number: int|None = None ) -> stream.Opus: ''' Convert a multi-work stream into one or more complete works packed into an Opus Stream. If a `number` argument is given, and a work is defined by that number, an Opus containing just that single work is returned. * Changed in v11: always returns an Opus even if only a single number is given. ''' if inputM21 is None: opus = stream.Opus() else: opus = inputM21 # environLocal.printDebug(['abcToStreamOpus: got number', number]) # returns a dictionary of numerical key if abcHandler.definesReferenceNumbers(): abcDict = abcHandler.splitByReferenceNumber() if number is not None and number in abcDict: # get the single requested work, but still wrap it in an Opus opus.append(abcToStreamScore(abcDict[number])) else: # build entire opus into an opus stream scoreList = [] for key, value in sorted(abcDict.items()): # do not need to set work number, as that will be gathered # with metadata in abcToStreamScore try: sc = abcToStreamScore(value) scoreList.append(sc) except IndexError: environLocal.warn(f'Failure for piece number {key}') for scoreDocument in scoreList: opus.coreAppend(scoreDocument, setActiveSite=False) opus.coreElementsChanged() else: # just return single entry in opus object opus.append(abcToStreamScore(abcHandler)) return opus
@typing.overload def reBar(music21Part: stream.Part, *, inPlace: typing.Literal[True]) -> None: ... @typing.overload def reBar(music21Part: stream.Part, *, inPlace: typing.Literal[False] = False) -> stream.Part: ...
[docs] def reBar(music21Part: stream.Part, *, inPlace: bool = False) -> stream.Part|None: # noinspection PyShadowingNames,SpellCheckingInspection ''' Re-bar overflow measures using the last known time signature. >>> irl2 = corpus.parse('irl', number=2) >>> irl2.metadata.title 'Aililiu na Gamhna, S.35' >>> music21Part = irl2[1] The whole part is in 2/4 time, but there are some measures expressed in 4/4 time without an explicit time signature change, an error in abc parsing due to the omission of barlines. The method will split those measures such that they conform to the last time signature, in this case 2/4. The default is to reBar in place. The measure numbers are updated accordingly. (NOTE: reBar is called automatically in abcToStreamPart, hence not demonstrated below.) The key signature and clef are assumed to be the same in the second measure after the split, so both are omitted. If the time signature is not the same in the second measure, the new time signature is indicated, and the measure following returns to the last time signature, except in the case that a new time signature is indicated. >>> music21Part.measure(15).show('text') {0.0} <music21.note.Note A> {1.0} <music21.note.Note A> >>> music21Part.measure(16).show('text') {0.0} <music21.note.Note A> {0.5} <music21.note.Note B-> {1.0} <music21.note.Note A> {1.5} <music21.note.Note G> An example where the time signature wouldn't be the same. This score is mistakenly marked as 4/4, but has some measures that are longer. >>> irl15 = corpus.parse('irl', number=15) >>> irl15.metadata.title 'Esternowe, S. 60' >>> music21Part2 = irl15.parts.first() # 4/4 time signature >>> music21Part2.measure(1).show('text') {0.0} <music21.note.Note C> {1.0} <music21.note.Note A> {1.5} <music21.note.Note G> {2.0} <music21.note.Note E> {2.5} <music21.note.Note G> >>> music21Part2.measure(1)[-1].duration.quarterLength 1.5 >>> music21Part2.measure(2).show('text') {0.0} <music21.meter.TimeSignature 1/8> {0.0} <music21.note.Note E> * Changed in v5: inPlace is False by default, and a keyword only argument. ''' if not inPlace: music21Part = copy.deepcopy(music21Part) lastTimeSignature = None measureNumberOffset = 0 # amount to shift current measure numbers allMeasures = music21Part.getElementsByClass(stream.Measure) for measureIndex in range(len(allMeasures)): music21Measure = allMeasures[measureIndex] if music21Measure.timeSignature is not None: lastTimeSignature = music21Measure.timeSignature if lastTimeSignature is None: raise ABCTranslateException('No time signature found in this Part') tsEnd = lastTimeSignature.barDuration.quarterLength mEnd = common.opFrac(music21Measure.highestTime) music21Measure.number += measureNumberOffset if mEnd > tsEnd: m1, m2 = music21Measure.splitAtQuarterLength(tsEnd) m2.timeSignature = None if lastTimeSignature.barDuration.quarterLength != m2.highestTime: try: m2.timeSignature = m2.bestTimeSignature() except exceptions21.StreamException as e: raise ABCTranslateException( f'Problem with measure {music21Measure.number} ({music21Measure!r}): {e}') if measureIndex != len(allMeasures) - 1: if allMeasures[measureIndex + 1].timeSignature is None: allMeasures[measureIndex + 1].timeSignature = lastTimeSignature m2.keySignature = None # suppress the key signature m2.clef = None # suppress the clef m2.number = m1.number + 1 measureNumberOffset += 1 music21Part.insert(common.opFrac(m1.offset + m1.highestTime), m2) # elif ((mEnd + music21Measure.paddingLeft) < tsEnd # and measureIndex != len(allMeasures) - 1): # The first and last measures are allowed to be incomplete # music21Measure.timeSignature = music21Measure.bestTimeSignature() # if allMeasures[measureIndex + 1].timeSignature is None: # allMeasures[measureIndex + 1].timeSignature = lastTimeSignature # if not inPlace: return music21Part
class ABCTranslateException(exceptions21.Music21Exception): pass # ------------------------------------------------------------------------------ class Test(unittest.TestCase): def testBasic(self): from music21 import abcFormat # from music21.abcFormat import testFiles # noinspection SpellCheckingInspection for tf in [ # testFiles.fyrareprisarn, # testFiles.mysteryReel, # testFiles.aleIsDear, # testFiles.testPrimitive, # testFiles.fullRiggedShip, # testFiles.kitchGirl, # testFiles.morrisonsJig, # testFiles.hectorTheHero, # testFiles.williamAndNancy, # testFiles.theAleWifesDaughter, # testFiles.theBeggerBoy, # testFiles.theAleWifesDaughter, # testFiles.draughtOfAle, # testFiles.testPrimitiveTuplet, # testFiles.testPrimitivePolyphonic, ]: af = abcFormat.ABCFile() ah = af.readstr(tf) # return handler, processes tokens s = abcToStreamScore(ah) s.show() # s.show('midi') def testGetMetaData(self): ''' NB -- only title is checked. not meter or key ''' from music21 import abcFormat from music21.abcFormat import testFiles for (tf, titleEncoded, unused_meterEncoded, unused_keyEncoded) in [ (testFiles.fyrareprisarn, 'Fyrareprisarn', '3/4', 'F'), (testFiles.mysteryReel, 'Mystery Reel', 'C|', 'G'), (testFiles.aleIsDear, 'The Ale is Dear', '4/4', 'D', ), (testFiles.kitchGirl, 'Kitchen Girl', '4/4', 'D'), (testFiles.williamAndNancy, 'William and Nancy', '6/8', 'G'), ]: af = abcFormat.ABCFile() ah = af.readstr(tf) # returns an ABCHandler object s = abcToStreamScore(ah) self.assertEqual(s.metadata.title, titleEncoded) def testChords(self): from music21 import abcFormat from music21.abcFormat import testFiles tf = testFiles.aleIsDear af = abcFormat.ABCFile() s = abcToStreamScore(af.readstr(tf)) # s.show() self.assertEqual(len(s.parts), 2) self.assertEqual(len(s.parts[0].flatten().notesAndRests), 111) self.assertEqual(len(s.parts[1].flatten().notesAndRests), 127) # chords are defined in second part here self.assertEqual(len(s.parts[1][chord.Chord]), 32) # check pitches in chords; sharps are applied due to key signature match = [p.nameWithOctave for p in s.parts[1].flatten().getElementsByClass( chord.Chord)[4].pitches] self.assertEqual(match, ['F#4', 'D4', 'B3']) match = [p.nameWithOctave for p in s.parts[1].flatten().getElementsByClass( chord.Chord)[3].pitches] self.assertEqual(match, ['E4', 'C#4', 'A3']) # s.show() # s.show('midi') def testMultiVoice(self): from music21 import abcFormat from music21.abcFormat import testFiles tf = testFiles.testPrimitivePolyphonic af = abcFormat.ABCFile() s = abcToStreamScore(af.readstr(tf)) self.assertEqual(len(s.parts), 3) # must flatten b/c there are measures self.assertEqual(len(s.parts[0].flatten().notesAndRests), 6) self.assertEqual(len(s.parts[1].flatten().notesAndRests), 17) self.assertEqual(len(s.parts[2].flatten().notesAndRests), 6) # s.show() # s.show('midi') def testTuplets(self): from music21 import abcFormat from music21.abcFormat import testFiles tf = testFiles.testPrimitiveTuplet af = abcFormat.ABCFile() s = abcToStreamScore(af.readstr(tf)) match = [] # match strings for better comparison for n in s.flatten().notesAndRests: match.append(n.quarterLength) shouldFind = [ 1 / 3, 1 / 3, 1 / 3, 1 / 5, 1 / 5, 1 / 5, 1 / 5, 1 / 5, 1 / 6, 1 / 6, 1 / 6, 1 / 6, 1 / 6, 1 / 6, 1 / 7, 1 / 7, 1 / 7, 1 / 7, 1 / 7, 1 / 7, 1 / 7, 2 / 3, 2 / 3, 2 / 3, 2 / 3, 2 / 3, 2 / 3, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 1 / 12, 2 ] self.assertEqual(match, [common.opFrac(x) for x in shouldFind]) def testAnacrusisPadding(self): from music21 import abcFormat from music21.abcFormat import testFiles # 2 quarter pickup in 3/4 ah = abcFormat.ABCHandler() ah.process(testFiles.hectorTheHero) s = abcToStreamScore(ah) m1 = s.parts[0].getElementsByClass(stream.Measure).first() # s.show() # ts is 3/4 self.assertEqual(m1.barDuration.quarterLength, 3.0) # filled with two quarter notes self.assertEqual(m1.duration.quarterLength, 2.0) # m1.show('t') # notes are shown as being on beat 2 and 3 # environLocal.printDebug(['m1.notesAndRests.activeSite', m1.notesAndRests.activeSite]) # environLocal.printDebug(['m1.notesAndRests[0].activeSite', # m1.notesAndRests[0].activeSite]) # self.assertEqual(m1.notesAndRests.activeSite) n0 = m1.notesAndRests[0] n1 = m1.notesAndRests[1] self.assertEqual(n0.getOffsetBySite(m1) + m1.paddingLeft, 1.0) self.assertEqual(m1.notesAndRests[0].beat, 2.0) self.assertEqual(n1.getOffsetBySite(m1) + m1.paddingLeft, 2.0) self.assertEqual(m1.notesAndRests[1].beat, 3.0) # two 16th pickup in 4/4 ah = abcFormat.ABCHandler() ah.process(testFiles.theAleWifesDaughter) s = abcToStreamScore(ah) m1 = s.parts[0].getElementsByClass(stream.Measure).first() # ts is 3/4 self.assertEqual(m1.barDuration.quarterLength, 4.0) # filled with two 16th self.assertEqual(m1.duration.quarterLength, 0.5) # notes are shown as being on beat 2 and 3 n0 = m1.notesAndRests[0] n1 = m1.notesAndRests[1] self.assertEqual(n0.getOffsetBySite(m1) + m1.paddingLeft, 3.5) self.assertEqual(m1.notesAndRests[0].beat, 4.5) self.assertEqual(n1.getOffsetBySite(m1) + m1.paddingLeft, 3.75) self.assertEqual(m1.notesAndRests[1].beat, 4.75) def testOpusImport(self): from music21 import corpus from music21 import abcFormat # replace w/ ballad80, smaller or erk5 fp = corpus.getWork('essenFolksong/teste') self.assertEqual(fp.name, 'teste.abc') self.assertEqual(fp.parent.name, 'essenFolksong') af = abcFormat.ABCFile() af.open(fp) # return handler, processes tokens ah = af.read() af.close() op = abcToStreamOpus(ah) # op.scores[3].show() self.assertEqual(len(op), 8) def testLyrics(self): # TODO(msc) -- test better from music21 import abcFormat from music21.abcFormat import testFiles tf = testFiles.sicutRosa af = abcFormat.ABCFile() s = abcToStreamScore(af.readstr(tf)) assert s is not None # s.show() # self.assertEqual(len(s.parts), 3) # self.assertEqual(len(s.parts[0].notesAndRests), 6) # self.assertEqual(len(s.parts[1].notesAndRests), 20) # self.assertEqual(len(s.parts[2].notesAndRests), 6) # s.show() # s.show('midi') def testMultiWorkImported(self): from music21 import corpus # defines multiple works, will return an opus o = corpus.parse('josquin/milleRegrets') self.assertEqual(len(o.scores), 4) # each score in the opus is a Stream that contains a Part and metadata p1 = o.getScoreByNumber(1).parts[0] self.assertEqual(p1.offset, 0.0) self.assertEqual(len(p1.flatten().notesAndRests), 90) p2 = o.getScoreByNumber(2).parts[0] self.assertEqual(p2.offset, 0.0) self.assertEqual(len(p2.flatten().notesAndRests), 80) p3 = o.getScoreByNumber(3).parts[0] self.assertEqual(p3.offset, 0.0) self.assertEqual(len(p3.flatten().notesAndRests), 86) p4 = o.getScoreByNumber(4).parts[0] self.assertEqual(p4.offset, 0.0) self.assertEqual(len(p4.flatten().notesAndRests), 78) sMerged = o.mergeScores() self.assertEqual(sMerged.metadata.title, 'Mille regrets') self.assertEqual(sMerged.metadata.composer, 'Josquin des Prez') self.assertEqual(len(sMerged.parts), 4) self.assertEqual(sMerged.parts[0][clef.Clef].first().sign, 'G') self.assertEqual(sMerged.parts[1][clef.Clef].first().sign, 'G') self.assertEqual(sMerged.parts[2][clef.Clef].first().sign, 'G') self.assertEqual(sMerged.parts[2][clef.Clef].first().octaveChange, -1) self.assertEqual(sMerged.parts[3][clef.Clef].first().sign, 'F') # sMerged.show() def testChordSymbols(self): from music21 import corpus from music21 import pitch # noinspection SpellCheckingInspection o = corpus.parse('nottingham-dataset/reelsa-c') self.assertEqual(len(o.scores), 2) # each score in the opus is a Stream that contains a Part and metadata p1 = o.getScoreByNumber(81).parts[0] self.assertEqual(p1.offset, 0.0) self.assertEqual(len(p1.flatten().notesAndRests), 77) self.assertEqual(len(list(p1.flatten().getElementsByClass(harmony.ChordSymbol))), 25) # Am/C self.assertEqual(list(p1.flatten().getElementsByClass(harmony.ChordSymbol))[7].root(), pitch.Pitch('A3')) self.assertEqual(list(p1.flatten().getElementsByClass(harmony.ChordSymbol))[7].bass(), pitch.Pitch('C3')) # G7/B self.assertEqual(list(p1.flatten().getElementsByClass(harmony.ChordSymbol))[14].root(), pitch.Pitch('G3')) self.assertEqual(list(p1.flatten().getElementsByClass(harmony.ChordSymbol))[14].bass(), pitch.Pitch('B2')) def testNoChord(self): from music21 import converter target_str = ''' T: No Chords M: 4/4 L: 1/1 K: C [| "C" C | "NC" C | "C" C | "N.C." C | "C" C | "No Chord" C | "C" C | "None" C | "C" C | "Other" C |] ''' score = converter.parse(target_str, format='abc') self.assertEqual(len(score[harmony.ChordSymbol]), 9) self.assertEqual(len(score[harmony.NoChord]), 4) score = harmony.realizeChordSymbolDurations(score) self.assertEqual(8, score.getElementsByClass(harmony.ChordSymbol) .last().quarterLength) self.assertEqual(4, score.getElementsByClass(harmony.ChordSymbol) .first().quarterLength) def testAbcKeyImport(self): from music21 import abcFormat # sharps major = ['C', 'G', 'D', 'A', 'E', 'B', 'F#', 'C#'] minor = ['Am', 'Em', 'Bm', 'F#m', 'C#m', 'G#m', 'D#m', 'A#m'] for n, (majName, minName) in enumerate(zip(major, minor)): am = abcFormat.ABCMetadata('K:' + majName) am.preParse() ks_major = am.getKeySignatureObject() am = abcFormat.ABCMetadata('K:' + minName) am.preParse() ks_minor = am.getKeySignatureObject() self.assertEqual(n, ks_major.sharps) self.assertEqual(n, ks_minor.sharps) self.assertEqual('major', ks_major.mode) self.assertEqual('minor', ks_minor.mode) # flats major = ['C', 'F', 'Bb', 'Eb', 'Ab', 'Db', 'Gb', 'Cb'] minor = ['Am', 'Dm', 'Gm', 'Cm', 'Fm', 'Bbm', 'Ebm', 'Abm'] for n, (majName, minName) in enumerate(zip(major, minor)): am = abcFormat.ABCMetadata('K:' + majName) am.preParse() ks_major = am.getKeySignatureObject() am = abcFormat.ABCMetadata('K:' + minName) am.preParse() ks_minor = am.getKeySignatureObject() self.assertEqual(-1 * n, ks_major.sharps) self.assertEqual(-1 * n, ks_minor.sharps) self.assertEqual('major', ks_major.mode) self.assertEqual('minor', ks_minor.mode) # noinspection SpellCheckingInspection def testLocaleOfCompositionImport(self): from music21 import corpus # defines multiple works, will return an opus o = corpus.parse('essenFolksong/teste') self.assertEqual(len(o.scores), 8) s = o.getScoreByNumber(4) self.assertEqual(s.metadata.localeOfComposition, 'Asien, Ostasien, China, Sichuan') s = o.getScoreByNumber(7) self.assertEqual(s.metadata.localeOfComposition, 'Amerika, Mittelamerika, Mexiko') def testRepeatBracketsA(self): from music21.abcFormat import testFiles from music21 import converter s = converter.parse(testFiles.morrisonsJig) # s.show() # one start, one end # s.parts[0].show('t') self.assertEqual(len(s[bar.Repeat]), 2) # s.show() # this has a 1 note pickup # has three repeat bars; first one is implied s = converter.parse(testFiles.draughtOfAle) self.assertEqual(len(s[bar.Repeat]), 3) self.assertEqual(s[note.Note].first().pitch.nameWithOctave, 'D4') # new problem case: s = converter.parse(testFiles.hectorTheHero) # first measure has 2 pickup notes self.assertEqual(len(s.parts.first().getElementsByClass(stream.Measure).first().notes), 2) def testRepeatBracketsB(self): from music21.abcFormat import testFiles from music21 import converter from music21 import corpus s = converter.parse(testFiles.morrisonsJig) # TODO: get self.assertEqual(len(s[spanner.RepeatBracket]), 2) # s.show() # four repeat brackets here; 2 at beginning, 2 at end s = converter.parse(testFiles.hectorTheHero) self.assertEqual(len(s[spanner.RepeatBracket]), 4) s = corpus.parse('JollyTinkersReel') self.assertEqual(len(s[spanner.RepeatBracket]), 4) def testMetronomeMarkA(self): from music21.abcFormat import testFiles from music21 import converter s = converter.parse(testFiles.fullRiggedShip) mmStream = s.flatten().getElementsByClass(tempo.TempoIndication) self.assertEqual(len(mmStream), 1) self.assertEqual(str(mmStream[0]), '<music21.tempo.MetronomeMark Quarter=100>') s = converter.parse(testFiles.aleIsDear) mmStream = s.flatten().getElementsByClass(tempo.TempoIndication) # this is a two-part pieces, and this is being added for each part # not sure if this is a problem self.assertEqual(len(mmStream), 2) self.assertEqual(str(mmStream[0]), '<music21.tempo.MetronomeMark Quarter=211>') s = converter.parse(testFiles.theBeggerBoy) mmStream = s[tempo.TempoIndication] # this is a two-part pieces, and this is being added for each part # not sure if this is a problem self.assertEqual(len(mmStream), 1) self.assertEqual(str(mmStream[0]), '<music21.tempo.MetronomeMark maestoso Quarter=90>') # s.show() def testTranslateA(self): # this tests a few files in this collection, some of which are hard to # parse from music21 import corpus # noinspection SpellCheckingInspection for fn in ( 'ToCashellImGoingJig.abc', 'SundayIsMyWeddingDayJig.abc', 'SinkHimDoddieHighlandFling.abc', 'RandyWifeOfGreenlawReel.abc', 'PassionFlowerHornpipe.abc', 'NightingaleClog.abc', 'MountainRangerHornpipe.abc', 'LadiesPandelettsReel.abc', 'JauntingCarHornpipe.abc', 'GoodMorrowToYourNightCapJig.abc', 'ChandlersHornpipe.abc', 'AlistairMaclalastairStrathspey.abc', ): s = corpus.parse(fn) assert s is not None # s.show() def testCleanFlat(self): from music21 import pitch cs = harmony.ChordSymbol(root='eb', bass='bb', kind='dominant') self.assertEqual(cs.bass(), pitch.Pitch('B-2')) self.assertIs(cs.pitches[0], cs.bass()) cs = harmony.ChordSymbol('e-7/b-') self.assertEqual(cs.root(), pitch.Pitch('E-3')) self.assertEqual(cs.bass(), pitch.Pitch('B-2')) self.assertEqual(cs.pitches[0], pitch.Pitch('B-2')) # common.cleanedFlatNotation() shouldn't be called by # the following calls, which what is being tested here: cs = harmony.ChordSymbol('b-3') self.assertEqual(cs.root(), pitch.Pitch('b-3')) self.assertEqual(cs.pitches[0], pitch.Pitch('B-3')) self.assertEqual(cs.pitches[1], pitch.Pitch('D4')) cs = harmony.ChordSymbol('bb3') # B, not B-flat self.assertEqual(cs.root(), pitch.Pitch('b2')) # b3 alteration applied to B major triad self.assertEqual(cs.pitches[0], pitch.Pitch('B2')) self.assertEqual(cs.pitches[1], pitch.Pitch('D3')) self.assertEqual(cs.pitches[2], pitch.Pitch('F#3')) def xtestTranslateB(self): ''' Dylan -- this could be too slow to make it a test! Numbers 637 and 749 fail ''' from music21 import corpus for fn in ['airdsAirs/book4.abc']: s = corpus.parse(fn) assert s is not None # s.show() def testTranslateBrokenDuration(self): from music21 import corpus unused = corpus.parse('han2.abc', number=445) def testTiesTranslate(self): from music21 import converter notes = converter.parse('L:1/8\na-a-a', format='abc') ties = [n.tie.type for n in notes.flatten().notesAndRests] self.assertListEqual(ties, ['start', 'continue', 'stop']) def xtestMergeScores(self): from music21 import corpus unused = corpus.parse('josquin/laDeplorationDeLaMorteDeJohannesOckeghem') # this was getting incorrect Clefs if __name__ == '__main__': import music21 music21.mainTest(Test)