def spell_time_absolute(self, time):
"""\
Convert a time expression into words.
:param time: The 12hr numerical time value in a string, e.g. '08:05:pm'
:return: time string with all numerals written out as words
"""
period = ""
if len(time.split(':')) > 2:
hours, mins, period = time.split(':')[0:3]
elif len(time.split(':')) > 1:
hours, mins, = time.split(':')[0:2]
else:
return time # 'now' and similar
hours = int(hours)
mins = int(mins)
# hack for TTS
period = "P M" if period == "PM" else period
period = "A M" if period == "AM" else period
time_str = [word_for_number(hours)] # always start with an hour
if mins == 0:
return ' '.join(time_str + [period])
if mins < 9:
time_str.append('o') # support one-o-one am
return ' '.join(time_str + [word_for_number(mins), period])
def __init__(self, *args, **kwargs):
super(PTIENSLUPreprocessing, self).__init__(*args, **kwargs)
num_norms = []
for num in xrange(60):
num_norms.append(([unicode(num)], word_for_number(num).split()))
self.text_normalization_mapping += num_norms
# map ordinal numbers for streets and avenues 1st - 200th
ord_norms = []
suffixes = ['th', 'st', 'nd', 'rd', 'th', 'th', 'th', 'th', 'th', 'th']
for ord in xrange(1, 200):
suffix = suffixes[ord % 10] if ord < 10 or ord > 20 else 'th'
ord_form = str(ord) + suffix
ord_norms.append(
([unicode(ord_form)], word_for_number(ord,
ordinary=True).split()))
self.text_normalization_mapping += ord_norms
self.text_normalization_mapping += [
(["i'm"], ['i am']),
(["im"], ['i am']),
(["it'll"], ['it will']),
(["i'll"], ['i will']),
(["that's"], ['that is']),
(["don't"], ['do not']),
(["doesn't"], ['does not']),
]
def __init__(self, *args, **kwargs):
super(PTIENSLUPreprocessing, self).__init__(*args, **kwargs)
num_norms = []
for num in xrange(60):
num_norms.append(([unicode(num)], word_for_number(num).split()))
self.text_normalization_mapping += num_norms
# map ordinal numbers for streets and avenues 1st - 200th
ord_norms = []
suffixes = ['th', 'st', 'nd', 'rd', 'th', 'th', 'th', 'th', 'th', 'th']
for ord in xrange(1, 200):
suffix = suffixes[ord % 10] if ord < 10 or ord > 20 else 'th'
ord_form = str(ord) + suffix
ord_norms.append(([unicode(ord_form)], word_for_number(ord, ordinary=True).split()))
self.text_normalization_mapping += ord_norms
self.text_normalization_mapping += [
(["i'm"], ['i am']),
(["im"], ['i am']),
(["it'll"], ['it will']),
(["i'll"], ['i will']),
(["that's"], ['that is']),
(["don't"], ['do not']),
(["doesn't"], ['does not']),
]
def spell_time_relative(self, time):
"""\
Convert a time expression into words.
:param time: Numerical time value in a string, e.g. '8:05'
:return: time string with all numerals written out as words 0:15 will generate '15 minutes' and not '0 hours and \
15 minutes'.
"""
if ':' not in time: # 'now' and similar
return time
hours, mins = map(int, time.split(':'))
time_str = []
if hours is not 0:
hr_id = 'hour' if hours is 1 else 'hours'
hours = word_for_number(hours)
time_str.extend((hours, hr_id))
if mins == 0 and hours != 0:
return ' '.join(time_str)
if time_str:
time_str.append('and')
min_id = 'minute' if mins == 1 else 'minutes'
mins = word_for_number(mins)
return ' '.join(time_str + [mins, min_id])
def spell_time_absolute(self, time):
"""\
Convert a time expression into words.
:param time: The 12hr numerical time value in a string, e.g. '08:05:pm'
:return: time string with all numerals written out as words
"""
period = ""
if len(time.split(':')) > 2:
hours, mins, period = time.split(':')[0:3]
elif len(time.split(':')) > 1:
hours, mins, = time.split(':')[0:2]
else:
return time # 'now' and similar
hours = int(hours)
mins = int(mins)
# hack for TTS
period = "P M" if period == "PM" else period
period = "A M" if period == "AM" else period
time_str = [word_for_number(hours)] # always start with an hour
if mins == 0:
return ' '.join(time_str + [period])
if mins < 9:
time_str.append('o') # support one-o-one am
return ' '.join(time_str + [word_for_number(mins), period])
def spell_time_relative(self, time):
"""\
Convert a time expression into words.
:param time: Numerical time value in a string, e.g. '8:05'
:return: time string with all numerals written out as words 0:15 will generate '15 minutes' and not '0 hours and \
15 minutes'.
"""
if ':' not in time: # 'now' and similar
return time
hours, mins = map(int, time.split(':'))
time_str = []
if hours is not 0:
hr_id = 'hour' if hours is 1 else 'hours'
hours = word_for_number(hours)
time_str.extend((hours, hr_id))
if mins == 0 and hours != 0:
return ' '.join(time_str)
if time_str:
time_str.append('and')
min_id = 'minute' if mins == 1 else 'minutes'
mins = word_for_number(mins)
return ' '.join(time_str + [mins, min_id])
def spell_temperature(self, value, interval):
"""Convert a temperature expression into words (assuming nominative).
:param value: Temperature value (whole number in degrees as string), \
e.g. '1' or '-10'.
:param interval: Boolean indicating whether to treat this as a start \
of an interval, i.e. omit the degrees word.
:return: temperature expression as string
"""
ret = ''
value = int(value)
if value < 0:
ret += 'minus '
value = abs(value)
ret += word_for_number(value)
if not interval:
ret += ' degrees'
return ret
def spell_temperature(self, value, interval):
"""Convert a temperature expression into words (assuming nominative).
:param value: Temperature value (whole number in degrees as string), \
e.g. '1' or '-10'.
:param interval: Boolean indicating whether to treat this as a start \
of an interval, i.e. omit the degrees word.
:return: temperature expression as string
"""
ret = ''
value = int(value)
if value < 0:
ret += 'minus '
value = abs(value)
ret += word_for_number(value)
if not interval:
ret += ' degrees'
return ret
def spell_number_range(start=0, end=10):
spelled = []
while end < start:
spelled.append(word_for_number(start))
start += 1
return spelled
def spell_number_range(start=0, end=10):
spelled = []
while end < start:
spelled.append(word_for_number(start))
start += 1
return spelled
