def alignDependingOnWithDuration(URIrecordingNoExt, whichSection, pathToComposition, withDuration, withSynthesis, evalLevel, params, usePersistentFiles, htkParser):
'''
call alignment method depending on whether duration or htk selected
'''
Phonetizer.initLookupTable(withSynthesis)
tokenLevelAlignedSuffix, phonemesAlignedSuffix = determineSuffix(withDuration, withSynthesis, evalLevel)
if withDuration:
alignmentErrors, detectedWordList, grTruthDurationWordList = alignOneChunk(URIrecordingNoExt, pathToComposition, whichSection, htkParser, params, evalLevel, usePersistentFiles)
else:
URIrecordingAnno = URIrecordingNoExt + ANNOTATION_EXT
URIrecordingWav = URIrecordingNoExt + AUDIO_EXTENSION
# new makamScore used
lyricsObj = loadLyrics(pathToComposition, whichSection)
lyrics = lyricsObj.__str__()
# in case we are at no-lyrics section
if not lyrics or lyrics =='_SAZ_':
logger.warn("skipping section {} with no lyrics ...".format(whichSection))
return [], [], [], []
outputHTKPhoneAlignedURI = Aligner.alignOnechunk(MODEL_URI, URIrecordingWav, lyrics.__str__(), URIrecordingAnno, '/tmp/', withSynthesis)
alignmentErrors = evalAlignmentError(URIrecordingAnno, outputHTKPhoneAlignedURI, evalLevel)
detectedWordList = outputHTKPhoneAlignedURI
grTruthDurationWordList = []
# store decoding results in a file FIXME: if with duration it is not mlf
detectedAlignedfileName = []
detectedAlignedfileName = tokenList2TabFile(detectedWordList, URIrecordingNoExt, tokenLevelAlignedSuffix)
return alignmentErrors, detectedWordList, grTruthDurationWordList, detectedAlignedfileName
def expandToPhonemes(self):
'''
make sure text has no whitespaces
'''
# METUtext = PhonetizerOlder.turkishScriptWord2METUScriptWord(self.text)
# phonemeIDs = PhonetizerOlder.grapheme2Phoneme(METUtext)
#
if not Phonetizer.lookupTable:
sys.exit("Phonetizer.lookupTable not defined. do Phonetizer.initlookupTable at beginning of all code")
self.phonemes = []
# instrument
if self.text == '_SAZ_':
self.phonemes.append(Phoneme('sil'))
# text from lyrics
else:
phonemeIDs = Phonetizer.grapheme2Phoneme(self.text)
for phonemeID in phonemeIDs:
self.phonemes.append(Phoneme(phonemeID))
if self.hasShortPauseAtEnd:
self.phonemes.append(Phoneme('sp'))
def loadMakamScore(pathToComposition):
'''
same as loadLyrics, but return MakamScore, so that all lyrics can be shown if needed
'''
Phonetizer.initLookupTable(False)
os.chdir(pathToComposition)
pathTotxt = os.path.join(pathToComposition, glob.glob("*.txt")[0])
listExtensions = ["sections.json", "sections.tsv", "sections.txt"]
sectionFiles = findFileByExtensions(pathToComposition, listExtensions)
sectionFile = sectionFiles[0]
pathToSectionTsv = os.path.join(pathToComposition, sectionFile)
makamScore = MakamScore(pathTotxt, pathToSectionTsv)
return makamScore
def loadMakamScore(pathToComposition):
'''
same as loadLyrics, but return MakamScore, so that all lyrics can be shown if needed
'''
Phonetizer.initLookupTable(False)
os.chdir(pathToComposition)
pathTotxt = os.path.join(pathToComposition, glob.glob("*.txt")[0])
listExtensions = [ "sections.json", "sections.tsv", "sections.txt"]
sectionFiles = findFileByExtensions(pathToComposition, listExtensions)
sectionFile = sectionFiles[0]
pathToSectionTsv = os.path.join(pathToComposition, sectionFile)
makamScore = MakamScore(pathTotxt, pathToSectionTsv )
return makamScore
def syllables2Lyrics(syllables):
listWords = []
for syllable in syllables:
# word of only one syllable
word, dummy = createWord([], syllable)
listWords.append(word)
Phonetizer.initLookupTable(True, 'phonemeMandarin2METUphonemeLookupTableSYNTH')
# load phonetic dict
Phonetizer.initPhoneticDict('syl2phn46.txt')
## 3) create lyrics
# here is called Syllable.expandToPhonemes.
lyrics = Lyrics(listWords)
return lyrics
def loadLyrics(pathToComposition, whichSection):
Phonetizer.initLookupTable(False)
os.chdir(pathToComposition)
pathTotxt = os.path.join(pathToComposition, glob.glob("*.txt")[0])
listExtensions = ["sections.json", "sections.tsv", "sections.txt"]
sectionFiles = findFileByExtensions(pathToComposition, listExtensions)
sectionFile = sectionFiles[0]
pathToSectionTsv = os.path.join(pathToComposition, sectionFile)
makamScore = MakamScore(pathTotxt, pathToSectionTsv)
# phoneme IDs
lyrics = makamScore.getLyricsForSection(whichSection)
return lyrics
def loadLyrics(pathToComposition, whichSection):
Phonetizer.initLookupTable(False)
os.chdir(pathToComposition)
pathTotxt = os.path.join(pathToComposition, glob.glob("*.txt")[0])
listExtensions = [ "sections.json", "sections.tsv", "sections.txt"]
sectionFiles = findFileByExtensions(pathToComposition, listExtensions)
sectionFile = sectionFiles[0]
pathToSectionTsv = os.path.join(pathToComposition, sectionFile)
makamScore = MakamScore(pathTotxt, pathToSectionTsv )
# phoneme IDs
lyrics = makamScore.getLyricsForSection(whichSection)
return lyrics
def expandToPhonemes(self):
'''
one-to-one function: PINYIN characters to turkish METU.
as intermediate step mandarin phoneset are used.
'''
######################
### pinyin to mandarin phonemeSet
if not Phonetizer.phoneticDict:
sys.exit("Phonetizer.phonetic Dict not defined. do Phonetizer.initPhoneticDict at beginning of all code")
self.phonemes = []
# notes for instrument
if self.text == 'REST':
# TODO: replace with other model instead of silence
self.phonemes.append(Phoneme('sil'))
# TODO: does sp at end of sp make sence?
self.phonemes.append(Phoneme('sp'))
return
if self.text not in Phonetizer.phoneticDict:
sys.exit("unknown syllable {} Please add it to the dict".format(self.text))
mandarinPhonemes = Phonetizer.phoneticDict[self.text]
####################
#### mandarin to turkish METU
if not Phonetizer.lookupTable:
sys.exit("Phonetizer.lookupTable not defined. do Phonetizer.initlookupTable at beginning of all code")
phonemeIDs = []
for ph in mandarinPhonemes:
phonemeIDs = Phonetizer.grapheme2phonemeList(ph, phonemeIDs)
#### create Phonemes as field
for phonemeID in phonemeIDs:
self.phonemes.append(Phoneme(phonemeID))
if self.hasShortPauseAtEnd:
self.phonemes.append(Phoneme('sp'))
syllablesInCurrWord.append(currSyllable)
beginIndex = beginIndex + 1
return listWords, beginIndex
#################################################################################
if __name__ == "__main__":
pathTxt= '/Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data/nihavent--sarki--aksak--bakmiyor_cesm-i--haci_arif_bey/nihavent--sarki--aksak--bakmiyor_cesm-i--haci_arif_bey.txt'
pathTsv= '/Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data/nihavent--sarki--aksak--bakmiyor_cesm-i--haci_arif_bey/nihavent--sarki--aksak--bakmiyor_cesm-i--haci_arif_bey.sections.tsv'
pathTxt= '/Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin.txt'
pathTsv= '/Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin.sections.tsv'
Phonetizer.initLookupTable(False)
symbTrParser = SymbTrParser(pathTxt, pathTsv)
symbTrParser.syllables2Lyrics()
# print them, put instrument as name
print symbTrParser.sectionLyrics[5]
print symbTrParser.sectionLyrics[0]
#restart counting
syllablesInCurrWord = []
else:
syllablesInCurrWord.append(currSyllable)
beginIndex = beginIndex + 1
return listWords, beginIndex
#################################################################################
if __name__ == "__main__":
pathTxt = '/Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data/nihavent--sarki--aksak--bakmiyor_cesm-i--haci_arif_bey/nihavent--sarki--aksak--bakmiyor_cesm-i--haci_arif_bey.txt'
pathTsv = '/Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data/nihavent--sarki--aksak--bakmiyor_cesm-i--haci_arif_bey/nihavent--sarki--aksak--bakmiyor_cesm-i--haci_arif_bey.sections.tsv'
pathTxt = '/Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin.txt'
pathTsv = '/Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin.sections.tsv'
Phonetizer.initLookupTable(False)
symbTrParser = SymbTrParser(pathTxt, pathTsv)
symbTrParser.syllables2Lyrics()
# print them, put instrument as name
print symbTrParser.sectionLyrics[5]
print symbTrParser.sectionLyrics[0]
def loadLyricsFromTextGridSentence(currSentence):
Phonetizer.initLookupTable(True, 'phonemeMandarin2METUphonemeLookupTableSYNTH')
syllables = currSentence[4]
lyrics = syllables2Lyrics(syllables)
return lyrics
pathHMMDuration = os.path.join(parentDir, 'AlignmentDuration')
if pathHMMDuration not in sys.path:
sys.path.append(pathHMMDuration)
from Phonetizer import Phonetizer
if __name__ == '__main__':
aria = 'dan-xipi_02'
aria = 'laosheng-xipi_02'
aria = 'laosheng-erhuang_04'
URI = '/Users/joro/Documents/Phd/UPF/arias_dev_01_t_70/'
MusicXmlURI = URI + aria + '_score.xml'
lyricsTextGrid = URI + aria + '.TextGrid'
# MusicXmlURI = 'dan-xipi_01_score.xml'
# lyricsTextGrid = 'dan-xipi_01.TextGrid'
musicXMLParser = MusicXMLParser(MusicXmlURI, lyricsTextGrid)
Phonetizer.initLookupTable(True,
'phonemeMandarin2METUphonemeLookupTableSYNTH')
for i in range(len(musicXMLParser.listSentences)):
print i, " ", musicXMLParser.getLyricsForSection(i)
# for syll in musicXMLParser.listSyllables:
# print syll
sys.path.append(pathHMMDuration)
from Phonetizer import Phonetizer
if __name__=='__main__':
aria = 'dan-xipi_02'
aria = 'laosheng-xipi_02'
aria = 'laosheng-erhuang_04'
URI = '/Users/joro/Documents/Phd/UPF/arias_dev_01_t_70/'
MusicXmlURI = URI + aria + '_score.xml'
lyricsTextGrid = URI + aria + '.TextGrid'
# MusicXmlURI = 'dan-xipi_01_score.xml'
# lyricsTextGrid = 'dan-xipi_01.TextGrid'
musicXMLParser = MusicXMLParser(MusicXmlURI, lyricsTextGrid)
Phonetizer.initLookupTable(True, 'phonemeMandarin2METUphonemeLookupTableSYNTH')
for i in range(len(musicXMLParser.listSentences)):
print i, " ", musicXMLParser.getLyricsForSection(i)
# for syll in musicXMLParser.listSyllables:
# print syll
