def __init__(self,
max_files=1000000,
file_lines=FILE_LINES,
batch_size=32,
shuffle=True):
'Initialization'
self.raw_dir = RAW_DIR
self.cache_dir = CACHE_DIR
self.raw_files = []
self.cache_files = []
files = []
for file in os.listdir(self.raw_dir):
files.append(file)
if len(files) >= max_files:
break
for file in files:
self.raw_files.append(self.raw_dir + '/' + file)
for file in files:
self.cache_files.append(self.cache_dir + '/' + file + '.pk')
# arange does this
self.indexes = [1] * len(self.raw_files)
for i in range(len(self.raw_files)):
self.indexes[i] = i
self.file_lines = file_lines
self.batch_size = batch_size
self.syll_mgr = syllables.syllables()
self.n_classes = self.syll_mgr.get_size()
self.shuffle = shuffle
self.on_epoch_end()
self.cmudict = cmudict.CMUDict()
def __init__(self,
dir=DIR,
file_lines=FILE_LINES,
num_lines=LINES,
batch_size=32,
shuffle=True):
'Initialization'
self.dir = dir
self.list_IDs = []
for file in os.listdir(dir):
self.list_IDs.append(dir + '/' + file)
self.indexes = [1] * len(self.list_IDs)
for i in range(len(self.list_IDs)):
self.indexes[i] = i
self.file_lines = file_lines
self.num_lines = num_lines
self.batch_size = batch_size
self.syll_mgr = syllables.syllables()
self.n_classes = self.syll_mgr.get_size()
self.shuffle = shuffle
self.on_epoch_end()
self.cmudict = cmudict.CMUDict()
from sets import Set
import sys
import cmudict
import tokens
cmudict = cmudict.CMUDict()
wordset = Set([])
for line in sys.stdin:
text = line.split("\t")[0]
words = tokens.tokenize(line)
words = tokens.fixtokens(words)
words = tokens.hyphen(words, cmudict.syll_dict)
for word in words:
wordset.add(word)
for word in sorted(wordset):
print(word)
next = list(sentence)
next.append(l)
yield next
recurse(syll_list, poss_array, 0)
#print(poss_array)
sentences = []
min_len = len(syll_list) + 1
for sentence in walktree(poss_array, []):
sentence = list(sentence)
if len(sentence) < min_len:
min_len = len(sentence)
sentences.append(" ".join(sentence))
out = []
for s in set(sentences):
if len(s.split(' ')) == min_len:
out.append(s)
return out
if __name__ == "__main__":
decoder = Decoder(cmudict.CMUDict())
for x in decoder.decode_sentence(['DH AH', 'S AH N', 'L IH T']):
print(x)
print(' ')
for x in decoder.decode_sentence([
'DH AH', 'S AH N', 'L IH T', 'AA', 'N IH NG', 'HH IY', 'V IH NG',
'OW', 'V ER', 'HH EH D'
]):
print(x)
# haiku gen 1 column
import sys
import snlp
import cmudict
nsyll = 5
cd = cmudict.CMUDict()
# TODO
# looking good except for sentences with hyphenated words.
# they don't get clipped, just keep going ??
num_lines = 0
num_pass = 0
num_nonword = 0
outf_5 = open("haiku_5.txt", "w")
outf_7 = open("haiku_7.txt", "w")
num_5 = 0
num_7 = 0
lastline = ''
for line in sys.stdin:
if line == lastline:
continue
lastline = line
line = line[:-1].lower()
if line[0] != '(':
continue
if len(stress) == len(meter):
fail = self.distance(stress, meter)
elif len(stress) + 1 == meter:
fail = self.distance('?' + stress, meter)
if fail > 1:
fail = self.distance(stress + '?', meter)
if fail < 2:
poss.append(name)
return poss
if __name__ == "__main__":
data = [["short!", ['1', '10']], ["long!", ['11001000000']],
["swap", ['01', '10', '0', '1', '0101']],
["correct", ['0101010101']], ["one", ['0111010101']],
["two!", ['0111010001']], ["first missing", ['101010101']],
["last missing", ['010101010']]]
meter = meter(cmudict.CMUDict(), {"meter": "01"}, "/def/null")
print("Strict meter:")
for test in data:
print("{0} -> {1}".format(test[0], meter.meter_strict(test[1])))
print("Broken meter:")
for test in data:
print("{0} -> {1}".format(test[0], meter.meter_loose(test[1])))
print(meter.possibles(['a', 'word'], {'a': 0, 'word': 0, 'word(2)': 0}))