def general_trans(self, text, filter_func,
normpunc=False, ligatures=False):
"""Transliaterates a word into IPA, filtering with filter_func
Args:
text (str): word to transcribe; unicode strings
filter_func (function): function for filtering segments; takes
a <segment, is_ipa> tuple and returns a
boolean.
normpunc (bool): normalize punctuation
ligatures (bool): use precomposed ligatures instead of
standard IPA
Returns:
unicode: IPA string, filtered by filter_func.
"""
text = unicode(text)
text = unicodedata.normalize('NFD', text.lower())
logging.debug('(after norm) text=' + repr(list(text)))
text = self.strip_diacritics.process(text)
logging.debug('(after strip) text=' + repr(list(text)))
if self.preproc:
text = self.preprocessor.process(text)
logging.debug('(after preproc) text=' + repr(list(text)))
tr_list = []
while text:
logging.debug('text=' + repr(list(text)))
m = self.regexp.match(text)
if m:
source = m.group(0)
try:
target = self.g2p[source][0]
except KeyError:
logging.debug("source = '{}'".format(source))
logging.debug("self.g2p[source] = '{}'"
.format(self.g2p[source]))
target = source
except IndexError:
logging.debug("self.g2p[source]={}".format(self.g2p[source]))
target = source
tr_list.append((target, True))
text = text[len(source):]
else:
tr_list.append((text[0], False))
self.nils[text[0]] += 2
text = text[1:]
text = ''.join([s for (s, _) in filter(filter_func, tr_list)])
if self.postproc:
text = self.postprocessor.process(text)
if ligatures or self.ligatures:
text = ligaturize(text)
if normpunc:
text = self.puncnorm.norm(text)
return unicodedata.normalize('NFC', text)
def general_trans(self, text, filter_func,
normpunc=False, ligatures=False):
"""Transliaterates a word into IPA, filtering with filter_func
Args:
text (str): word to transcribe; unicode strings
filter_func (function): function for filtering segments; takes
a <segment, is_ipa> tuple and returns a
boolean.
normpunc (bool): normalize punctuation
ligatures (bool): use precomposed ligatures instead of
standard IPA
Returns:
unicode: IPA string, filtered by filter_func.
"""
text = unicode(text)
text = self.strip_diacritics.process(text)
text = unicodedata.normalize('NFC', text.lower())
if self.preproc:
text = self.preprocessor.process(text)
tr_list = []
while text:
m = self.regexp.match(text)
if m:
source = m.group(0)
try:
target = self.g2p[source][0]
except KeyError:
logging.debug("source = '{}'".format(source))
logging.debug("self.g2p[source] = '{}'"
.format(self.g2p[source]))
target = source
tr_list.append((target, True))
text = text[len(source):]
else:
tr_list.append((text[0], False))
self.nils[text[0]] += 2
text = text[1:]
text = ''.join([s for (s, _) in filter(filter_func, tr_list)])
if self.postproc:
text = self.postprocessor.process(text)
if ligatures or self.ligatures:
text = ligaturize(text)
if normpunc:
text = self.puncnorm.norm(text)
return text
def transliterate(self, text, normpunc=False, ligatures=False, safe=True):
"""Transliterates/transcribes a word into IPA
Passes unmapped characters through to output unchanged.
Args:
word (str): word to transcribe; unicode string
normpunc (bool): normalize punctuation
ligatures (bool): use precomposed ligatures instead of standard IPA
Returns:
unicode: IPA string with unrecognized characters included
"""
text = unicode(text)
if safe and not self.regexp.search(text):
return text
text = self.strip_diacritics.process(text)
text = unicodedata.normalize('NFC', text.lower())
if self.preproc:
text = self.preprocessor.process(text)
tr_list = []
while text:
m = self.regexp.match(text)
if m:
from_seg = m.group(0)
try:
to_seg = self.g2p[from_seg][0]
except:
print("from_seg = {}".format(from_seg))
print("self.g2p[from_seg] = {}".format(self.g2p[from_seg]))
to_seg = from_seg
tr_list.append(to_seg)
text = text[len(from_seg):]
else:
tr_list.append(text[0])
self.nils[text[0]] += 1
text = text[1:]
text = ''.join(tr_list)
if self.postproc:
text = self.postprocessor.process(text)
if ligatures or self.ligatures:
text = ligaturize(text)
if normpunc:
text = self.puncnorm.norm(text)
return text
def transliterate(self, text, normpunc=False, ligatures=False):
"""Convert English text to IPA transcription
Args:
text (unicode): English text
normpunc (bool): if True, normalize punctuation downward
ligatures (bool): if True, use non-standard ligatures instead of
standard IPA
"""
text = unicodedata.normalize('NFC', text)
acc = []
for chunk in self.chunk_re.findall(text):
if self.letter_re.match(chunk):
acc.append(self.english_g2p(chunk))
else:
acc.append(chunk)
text = ''.join(acc)
text = self.puncnorm.norm(text) if normpunc else text
text = ligaturize(text) if (ligatures or self.ligatures) else text
return text
def transliterate(self, text, normpunc=False, ligatures=False):
"""Convert English text to IPA transcription
Args:
text (unicode): English text
normpunc (bool): if True, normalize punctuation downward
ligatures (bool): if True, use non-standard ligatures instead of
standard IPA
"""
text = unicodedata.normalize('NFC', text)
acc = []
for chunk in self.chunk_re.findall(text):
if self.letter_re.match(chunk):
acc.append(self.english_g2p(chunk))
else:
acc.append(chunk)
text = ''.join(acc)
text = self.puncnorm.norm(text) if normpunc else text
text = ligaturize(text) if ligatures else text
return text
