class Dataset:
# Dataset is a set of files associated with an audio file.
# The name of a dataset is same as the audio file.
# This class is an utility to manage this set of files
def __init__(self,name):
self.name = name
# A list of word(formed by phonetic)
self.charList = []
self._phoneticList = []
self._labelFile = None
def open(self):
charFile = open(self.name+".txt","r")
line = charFile.readline()
self.charList = [unicode(item,"utf-8") for item in line.split(" ")]
self._labelFile = LabelFile()
self._labelFile.open(self.name + "-label.txt")
self._samplingFile = SamplingFile()
self._samplingFile.open(self.name + "-sampling.csv")
self._createPhoneticList()
def phoneticList(self):
return self._phoneticList
def _createPhoneticList(self):
for row in self._labelFile:
sample = Sample()
sample.timestamp = row[0]
sample.phonetic = Ipa.simplifySymbol(row[2])
record = self._samplingFile.search(float(row[0]));
sample.zcr = float(record["ZCR"])
sample.variance = float(record["Spectrum Variance"])
sample.rms = float(record["RMS"])
sample.str = float(record["STE"])
self._phoneticList.append(sample)
def open(self):
charFile = open(self.name+".txt","r")
line = charFile.readline()
self.charList = [unicode(item,"utf-8") for item in line.split(" ")]
self._labelFile = LabelFile()
self._labelFile.open(self.name + "-label.txt")
self._samplingFile = SamplingFile()
self._samplingFile.open(self.name + "-sampling.csv")
self._createPhoneticList()
self._index = self._index+1
frame = self._sampling[self._index]
self._index = self._index + 1
return frame
try:
target = sys.argv[1]
except:
print "Usage: %s name_of_sample " % sys.argv[0]
exit(0)
table = SoundTable()
print "Reading sound.csv..."
table.open("sound.csv")
samplingFile = SamplingFile()
filename = target + "-sampling.csv"
print "Reading %s..." % filename
samplingFile.open(filename)
filename = target + ".txt"
print "Reading %s..." % filename
f = codecs.open(filename,"r","utf-8")
line = f.readline()
phonetics = [ Ipa(item.strip()) for item in line.split(" ") ]
filter = Filter()
filter.process(table,samplingFile,phonetics)
