def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('1')
f3.add_state('1a')
f3.add_state('1b')
f3.add_state('2')
f3.initial_state = '1'
# The soundex string will either need no padding in which case its final
# state is 1, or it will need 1 to 3 zeros and have final state 2
f3.set_final('1')
f3.set_final('2')
f3.add_arc('1', '2', (), ('000'))
f3.add_arc('1a', '2', (), ('00'))
f3.add_arc('1b', '2', (), ('0'))
for letter in string.letters:
f3.add_arc('1', '1', (letter), (letter))
for number in xrange(10):
f3.add_arc('1', '1a', (str(number)), (str(number)))
for number in xrange(10):
f3.add_arc('1a', '1b', (str(number)), (str(number)))
for number in xrange(10):
f3.add_arc('1b', '2', (str(number)), (str(number)))
return f3
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('st')
f3.add_state('L1')
f3.add_state('N1')
f3.add_state('N2')
f3.add_state('N3')
f3.add_state('P1')
f3.add_state('P2')
f3.add_state('P3')
f3.initial_state = 'st'
f3.set_final('N3')
f3.set_final('P3')
for letter in string.letters:
f3.add_arc('st', 'L1', (letter), (letter))
for number in xrange(10):
f3.add_arc('st', 'N1', (str(number)), (str(number)))
f3.add_arc('L1', 'N1', (str(number)), (str(number)))
f3.add_arc('N1', 'N2', (str(number)), (str(number)))
f3.add_arc('N2', 'N3', (str(number)), (str(number)))
f3.add_arc('L1', 'P1', (), ('0'))
f3.add_arc('N1', 'P2', (), ('0'))
f3.add_arc('N2', 'P3', (), ('0'))
f3.add_arc('P1', 'P2', (), ('0'))
f3.add_arc('P2', 'P3', (), ('0'))
return f3
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('1')
f3.add_state('2')
f3.add_state('3')
f3.add_state('4')
f3.add_state('5')
f3.add_state('6')
f3.add_state('7')
f3.add_state('8')
f3.initial_state = '1'
f3.set_final('5')
f3.set_final('8')
for letter in string.ascii_letters:
f3.add_arc('1', '2', letter, letter)
for num in string.digits:
f3.add_arc('1', '3', num, num)
f3.add_arc('2', '3', num, num)
f3.add_arc('3', '4', num, num)
f3.add_arc('4', '5', num, num)
f3.add_arc('2', '6', '', '0')
f3.add_arc('3', '7', '', '0')
f3.add_arc('4', '8', '', '0')
f3.add_arc('6', '7', '', '0')
f3.add_arc('7', '8', '', '0')
return f3
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
states = ['1', 'd1', 'd2', 'd3']
for state in states:
f2.add_state(state)
f2.initial_state = '1'
for state in ['d1', 'd2', 'd3']:
f2.set_final(state)
# Add the arcs
for letter in string.letters:
f2.add_arc('1', '1', (letter), (letter))
for index, state in enumerate(states):
if index > 0:
for n in range(10):
f2.add_arc(states[index-1], states[index], str(n), str(n))
for n in range(10):
f2.add_arc('d3', 'd3', str(n), ())
return f2
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('start')
f3.add_state('dig1')
f3.add_state('dig2')
f3.add_state('dig3')
f3.initial_state = 'start'
f3.set_final('dig3')
f3.add_arc('start', 'dig1', (), ('0'))
f3.add_arc('dig1', 'dig2', (), ('0'))
f3.add_arc('dig2', 'dig3', (), ('0'))
for letter in string.letters:
f3.add_arc('start', 'start', (letter), (letter))
for letter in ['1', '2', '3', '4', '5', '6']:
f3.add_arc('start', 'dig1', (letter), (letter))
f3.add_arc('dig1', 'dig2', (letter), (letter))
f3.add_arc('dig2', 'dig3', (letter), (letter))
return f3
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('start')
f2.add_state('0')
f2.add_state('1')
f2.add_state('2')
f2.add_state('3')
f2.initial_state = 'start'
f2.set_final('0')
f2.set_final('1')
f2.set_final('2')
f2.set_final('3')
# Add the arcs
for letter in string.letters:
f2.add_arc('start', '0', (letter), (letter))
for n in range(10):
f2.add_arc('start', '1', (str(n)), (str(n)))
f2.add_arc('0', '1', (str(n)), (str(n)))
f2.add_arc('1', '2', (str(n)), (str(n)))
f2.add_arc('2', '3', (str(n)), (str(n)))
f2.add_arc('3', '3', (str(n)), ())
return f2
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('1')
f3.add_state('2')
f3.add_state('3')
f3.add_state('4')
f3.add_state('5')
f3.initial_state = '1'
f3.set_final('4')
f3.set_final('1')
f3.add_arc('1', '4', (), '000')
f3.add_arc('2', '4', (), '00')
f3.add_arc('3', '4', (), '0')
for letter in string.letters:
f3.add_arc('1', '1', letter, letter)
for number in range(10):
f3.add_arc('1', '2', str(number), str(number))
for number in range(10):
f3.add_arc('2', '3', str(number), str(number))
for number in range(10):
f3.add_arc('3', '4', str(number), str(number))
for number in range(10):
f3.add_arc('4', '4', str(number), ())
return f3
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('s0')
f3.add_state('s1')
f3.add_state('s2')
f3.add_state('s3')
f3.add_state('s4')
f3.add_state('s5')
f3.add_state('s6')
f3.add_state('s7')
f3.add_state('s8')
f3.initial_state = 's0'
f3.set_final('s4')
f3.set_final('s7')
for letter in string.letters:
f3.add_arc('s0', 's1', (letter), (letter))
for digit in string.digits:
f3.add_arc('s1', 's5', (digit), (digit))
f3.add_arc('s5', 's6', (digit), (digit))
f3.add_arc('s6', 's7', (digit), (digit))
f3.add_arc('s0', 's8', (digit), (digit))
f3.add_arc('s1', 's2', (), ('0'))
f3.add_arc('s2', 's3', (), ('0'))
f3.add_arc('s3', 's4', (), ('0'))
f3.add_arc('s5', 's3', (), ('0'))
f3.add_arc('s6', 's4', (), ('0'))
f3.add_arc('s8', 's3', (), ('0'))
return f3
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('0')
f2.add_state('1')
f2.add_state('2')
f2.add_state('3')
f2.add_state('4')
f2.initial_state = '0'
f2.set_final('1')
f2.set_final('2')
f2.set_final('3')
f2.set_final('4')
# truncating extra digits.
numbers = '0123456789'
for letter in string.ascii_letters:
f2.add_arc('0', '1', letter, letter)
for number in numbers:
f2.add_arc('0', '2', number, number)
f2.add_arc('1', '2', number, number)
f2.add_arc('2', '3', number, number)
f2.add_arc('3', '4', number, number)
f2.add_arc('4', '4', number, '')
return f2
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('1')
f2.initial_state = '1'
for i in range(2, 6):
f2.add_state(str(i))
for i in range(2, 6):
f2.set_final(str(i))
for letter in string.letters:
f2.add_arc('1', '2', letter, letter)
f2.add_arc('2', '2', letter, letter)
for letter in ['1', '2', '3', '4', '5', '6']:
f2.add_arc('1', '3', letter, letter)
f2.add_arc('2', '3', letter, letter)
f2.add_arc('3', '4', letter, letter)
f2.add_arc('4', '5', letter, letter)
f2.add_arc('5', '5', letter, '')
return f2
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('1')
f2.add_state('2')
f2.add_state('3')
f2.add_state('4')
f2.initial_state = '1'
f2.set_final('1')
f2.set_final('2')
f2.set_final('3')
f2.set_final('4')
for letter in list(string.ascii_letters):
f2.add_arc('1', '1', letter, letter)
f2.add_arc('2', '1', letter, letter)
f2.add_arc('3', '1', letter, letter)
f2.add_arc('4', '1', letter, letter)
for digit in ['1', '2', '3', '4', '5', '6', '7', '8', '9', '0']:
f2.add_arc('1', '2', digit, digit)
f2.add_arc('2', '3', digit, digit)
f2.add_arc('3', '4', digit, digit)
f2.add_arc('4', '4', digit, '')
return f2
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('1')
f2.add_state('2')
f2.add_state('3')
f2.add_state('4')
f2.initial_state = '1'
f2.set_final('1')
f2.set_final('2')
f2.set_final('3')
f2.set_final('4')
# Adds letters from input string of 'A###0000'
for letter in string.letters:
f2.add_arc('1', '1', (letter), (letter))
# Adds numbers from first FST of range 0-9
for n in range(10):
f2.add_arc('1', '2', str(n), (str(n)))
f2.add_arc('2', '3', str(n), (str(n)))
f2.add_arc('3', '4', str(n), (str(n)))
f2.add_arc('4', '4', str(n), ())
return f2
def truncate_to_three_digits():
"""
A FST that will truncate a soundex string to three digits
"""
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('start')
f2.add_state('d1')
f2.add_state('d2')
f2.add_state('d3')
f2.add_state('end')
f2.initial_state = 'start'
f2.set_final('end')
# Add the arcs
for letter in string.letters:
f2.add_arc('start', 'd1', (letter), (letter))
f2.add_arc('d1', 'end', (), ())
f2.add_arc('d2', 'end', (), ())
f2.add_arc('d3', 'end', (), ())
for n in range(10):
f2.add_arc('d1', 'd2', (str(n)), (str(n)))
f2.add_arc('d2', 'd3', (str(n)), (str(n)))
f2.add_arc('d3', 'end', (str(n)), (str(n)))
f2.add_arc('end', 'end', (str(n)), ())
digits = ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']
for digit in digits:
f2.add_arc('start', 'd2', (digit), (digit))
return f2
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('1')
f3.add_state('2')
f3.add_state('3')
f3.add_state('4')
f3.add_state('5')
f3.add_state('6')
f3.initial_state = '1'
f3.set_final('6')
for letter in list(string.ascii_letters):
f3.add_arc('1', '1', letter, letter)
for digit in ['1', '2', '3', '4', '5', '6', '7', '8', '9', '0']:
f3.add_arc('1', '2', digit, digit)
f3.add_arc('2', '5', digit, digit)
f3.add_arc('5', '6', digit, digit)
f3.add_arc('1', '3', '', '0')
f3.add_arc('3', '4', '', '0')
f3.add_arc('4', '6', '', '0')
f3.add_arc('2', '4', '', '0')
f3.add_arc('5', '6', '', '0')
return f3
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('1')
f3.add_state('1a')
f3.add_state('1b')
f3.add_state('1c')
f3.add_state('1d')
f3.add_state('2')
f3.initial_state = '1'
f3.set_final('2')
for letter in string.letters:
f3.add_arc('1', '1', (letter), (letter))
for number in xrange(10):
f3.add_arc('1', '1a', (str(number)), (str(number)))
f3.add_arc('1a', '1b', (str(number)), (str(number)))
f3.add_arc('1b', '2', (str(number)), (str(number)))
f3.add_arc('1', '1c', (), ('0'))
f3.add_arc('1c', '1d', (), ('0'))
f3.add_arc('1d', '2', (), ('0'))
f3.add_arc('1a', '1d', (), ('0'))
f3.add_arc('1b', '2', (), ('0'))
return f3
def add_zero_padding():
# Third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('start')
f3.add_state('p1')
f3.add_state('p2')
f3.add_state('p3')
f3.add_state('end')
f3.initial_state = 'start'
f3.set_final('end')
f3.add_arc('start', 'p1', (), ())
for letter in string.letters:
f3.add_arc('start', 'p1', (letter), (letter))
f3.add_arc('p1', 'p2', (), ('0'))
f3.add_arc('p2', 'p3', (), ('0'))
f3.add_arc('p3', 'end', (), ('0'))
for number in xrange(10):
f3.add_arc('p1', 'p2', (str(number)), (str(number)))
f3.add_arc('p2', 'p3', (str(number)), (str(number)))
f3.add_arc('p3', 'end', (str(number)), (str(number)))
return f3
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('start2')
f3.add_state('char2')
f3.add_state('dig1')
f3.add_state('dig2')
f3.add_state('dig3')
f3.initial_state = 'start2'
f3.set_final('dig3')
f3.add_arc('char2', 'dig1', (), ('0'))
f3.add_arc('dig1', 'dig2', (), ('0'))
f3.add_arc('dig2', 'dig3', (), ('0'))
for letter in string.letters:
f3.add_arc('start2', 'char2', (letter), (letter))
for number in xrange(10):
f3.add_arc('start2', 'dig1', (str(number)), (str(number)))
f3.add_arc('char2', 'dig1', (str(number)), (str(number)))
f3.add_arc('dig1', 'dig2', (str(number)), (str(number)))
f3.add_arc('dig2', 'dig3', (str(number)), (str(number)))
f3.add_arc('dig3', 'dig3', (str(number)), ())
return f3
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
start_state = 'start'
letter_first = 'letter_first'
number_first = 'number_first'
numbers = list('0123456789')
# Initialization
f2 = FST('soundex-truncate')
f2.add_state(start_state)
f2.add_state(letter_first)
f2.add_state(number_first)
f2.set_final(number_first) #Don't think this would ever occur, but tests want it
f2.initial_state = start_state
for letter in string.ascii_letters:
f2.add_arc(start_state, letter_first, letter, letter)
for number in numbers:
f2.add_arc(start_state, number_first, str(number), str(number))
get_letter_number(f2, letter_first, numbers)
get_number_letter(f2, number_first, numbers)
return f2
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
# Indicate initial and final states
f3.add_state('0')
f3.add_state('1')
f3.add_state('1a')
f3.add_state('1b')
f3.add_state('1c')
f3.add_state('2')
f3.add_state('3a')
f3.add_state('3b')
f3.add_state('4')
f3.initial_state = '0'
f3.set_final('4')
# Add the arcs
for letter in string.letters:
f3.add_arc('0', '1', (letter), (letter))
for n in range(10):
f3.add_arc('1', '1a', (str(n)), (str(n)))
f3.add_arc('1a', '1b', (str(n)), (str(n)))
f3.add_arc('1b', '4', (str(n)), (str(n)))
f3.add_arc('1', '2', (str(n)), (str(n)))
f3.add_arc('1', '3a', (str(n)), (str(n)))
f3.add_arc('3a', '3b', (str(n)), (str(n)))
f3.add_arc('0', '1', (), ())
f3.add_arc('1', '4', (), ('000'))
f3.add_arc('2', '4', (), ('00'))
f3.add_arc('3b', '4', (), ('0'))
return f3
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
# Indicate initial and final states
f3.add_state('1')
f3.add_state('1a')
f3.add_state('1b')
f3.add_state('2')
f3.initial_state = '1'
f3.set_final('2')
# Add the arcs
for letter in string.letters:
f3.add_arc('1', '1', (letter), (letter))
for number in string.digits:
f3.add_arc('1', '1a', (number), (number))
f3.add_arc('1a', '1b', (number), (number))
f3.add_arc('1b', '2', (number), (number))
f3.add_arc('1', '2', (), ('000'))
f3.add_arc('1a', '2', (), ('00'))
f3.add_arc('1b', '2', (), ('0'))
return f3
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
states = ['1', 'd1', 'd2', 'd3']
for state in states:
f2.add_state(state)
f2.initial_state = '1'
for state in ['d1', 'd2', 'd3']:
f2.set_final(state)
# Add the arcs
for letter in string.letters:
f2.add_arc('1', '1', (letter), (letter))
for index, state in enumerate(states):
if index > 0:
for n in range(10):
f2.add_arc(states[index - 1], states[index], str(n), str(n))
for n in range(10):
f2.add_arc('d3', 'd3', str(n), ())
return f2
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('start')
f3.add_state('from_first_number')
f3.add_state('from_second_number')
f3.add_state('from_third_number')
f3.add_state('first_zero_append')
f3.add_state('second_zero_append')
f3.add_state('third_zero_append')
f3.initial_state = 'start'
f3.set_final('from_third_number')
f3.set_final('third_zero_append')
for letter in string.letters:
f3.add_arc('start', 'start', (letter), (letter))
f3.add_arc('start', 'first_zero_append', (), ('0'))
f3.add_arc('first_zero_append', 'second_zero_append', (), ('0'))
f3.add_arc('second_zero_append', 'third_zero_append', (), ('0'))
f3.add_arc('from_first_number', 'second_zero_append', (), ('0'))
f3.add_arc('from_second_number', 'third_zero_append', (), ('0'))
for number in xrange(10):
f3.add_arc('start', 'from_first_number', (str(number)), (str(number)))
f3.add_arc('from_first_number', 'from_second_number', (str(number)),
(str(number)))
f3.add_arc('from_second_number', 'from_third_number', (str(number)),
(str(number)))
return f3
def letters_to_numbers():
"""
Returns an FST that converts letters to numbers as specified by
the soundex algorithm
"""
# Let's define our first FST
f1 = FST('soundex-generate')
# Indicate that '1' is the initial state
f1.add_state('start')
f1.initial_state = 'start'
f1.add_state('0')
f1.set_final('0')
#===========================================================================
# for letter in string.ascii_letters:
# f1.add_arc('start', '0', (letter), (letter)) # for 1st letter in IP
#===========================================================================
removal_letters_set = {'a', 'e', 'i', 'o', 'u', 'h', 'w', 'y'}
for removeChar in list(removal_letters_set):
f1.add_arc('0', '0', (removeChar), ())
f1.add_arc('start', '0', (removeChar.lower()), (removeChar.lower()))
f1.add_arc('start', '0', (removeChar.upper()), (removeChar.upper()))
soundex_letter_lkp = [(['b', 'f', 'p', 'v'], '1'),
(['c', 'g', 'j', 'k', 'q', 's', 'x', 'z'], '2'),
(['d', 't'], '3'), (['l'], '4'), (['m', 'n'], '5'),
(['r'], '6')]
soundex_chars_set = set()
for charList, state in soundex_letter_lkp:
soundex_chars_set = soundex_chars_set.union(set(charList))
f1.add_state(state)
f1.set_final(state)
#build automata
for charList, state in soundex_letter_lkp:
for char in charList:
f1.add_arc('start', state, (char.lower()),
(char.lower())) #IF 1st jump to other soundex char
f1.add_arc('start', state, (char.upper()),
(char.upper())) #IF 1st jump to other soundex char
f1.add_arc('0', state, (char), (state)) #IF 1st jump to vowel
f1.add_arc(state, state, (char), ()) # self loop
for char in list(removal_letters_set): # for vowelsset
f1.add_arc(state, '0', (char), ())
for char in list(
soundex_chars_set.difference(set(charList))
): # any other char from different group will cause return to 0 state, with OP
for returnCharList, returnState in soundex_letter_lkp:
if char in returnCharList:
f1.add_arc(state, returnState, (char), (returnState))
return f1
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
f2.add_state('start')
f2.add_state('1')
f2.add_state('2')
f2.add_state('3')
f2.add_state('4')
f2.add_state('5')
f2.add_state('6')
f2.add_state('8')
f2.add_state('7')
f2.add_state('9')
f2.add_state('0')
f2.initial_state = 'start'
# Set all the final states
f2.set_final('2')
f2.set_final('1')
f2.set_final('3')
f2.set_final('4')
f2.set_final('5')
f2.set_final('6')
f2.set_final('7')
f2.set_final('8')
f2.set_final('9')
f2.set_final('0')
# Add the arcs
for letter in string.letters:
f2.add_arc('start', '0', (letter), (letter))
for letter in string.digits:
f2.add_arc('start', '5', (letter), (letter))
for letter in string.digits:
f2.add_arc('5', '6', (letter), (letter))
f2.add_arc('6', '7', (letter), (letter))
f2.add_arc('7', '7', (letter), ())
f2.add_arc('0', '1', (letter), (letter))
f2.add_arc('1', '2', (letter), (letter))
f2.add_arc('2', '3', (letter), (letter))
f2.add_arc('3', '3', (letter), ())
return f2
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('1')
f3.add_state('2')
f3.initial_state = '1'
f3.set_final('2')
return f3
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('1')
f3.add_state('2')
f3.initial_state = '1'
f3.set_final('2')
return f3
def french_count():
f = FST("french")
f.add_state("start")
f.initial_state = "start"
for ii in xrange(10):
f.add_arc("start", "start", str(ii), [kFRENCH_TRANS[ii]])
f.set_final("start")
return f
def french_count():
f = FST('french')
f.add_state('start')
f.initial_state = 'start'
for ii in xrange(10):
f.add_arc('start', 'start', str(ii), [kFRENCH_TRANS[ii]])
f.set_final('start')
return f
def french_count():
f = FST('french')
f.add_state('start')
f.initial_state = 'start'
for ii in xrange(10):
f.add_arc('start', 'start', [str(ii)], [kFRENCH_TRANS[ii]])
f.set_final('start')
return f
def letters_to_numbers():
"""
Returns an FST that converts letters to numbers as specified by
the soundex algorithm
"""
# Let's define our first FST
f1 = FST('soundex-generate')
# Indicate that '1' is the initial state
states = ['q1', 'q2', 'n1', 'n2', 'n3', 'n4', 'n5', 'n6']
for state in states:
f1.add_state(state)
f1.initial_state = 'q1'
# Set all the final states
for state in ['q2', 'n1', 'n2', 'n3', 'n4', 'n5', 'n6']:
f1.set_final(state)
# Add the rest of the arcs
for letter in string.ascii_lowercase:
f1.add_arc('q1', 'q2', (letter), (letter))
if letter in set('aehiouwy'):
for state in ['q2', 'n1', 'n2', 'n3', 'n4', 'n5', 'n6']:
f1.add_arc(state, state, (letter), ())
else:
if letter in set('bfpv'):
for state in ['q2', 'n2', 'n3', 'n4', 'n5', 'n6']:
f1.add_arc(state, 'n1', (letter), ('1'))
f1.add_arc('n1', 'n1', (letter), ())
elif letter in set('cgjkqsxz'):
for state in ['q2', 'n1', 'n3', 'n4', 'n5', 'n6']:
f1.add_arc(state, 'n2', (letter), ('2'))
f1.add_arc('n2', 'n2', (letter), ())
elif letter in set('dt'):
for state in ['q2', 'n1', 'n2', 'n4', 'n5', 'n6']:
f1.add_arc(state, 'n3', (letter), ('3'))
f1.add_arc('n3', 'n3', (letter), ())
elif letter in set('l'):
for state in ['q2', 'n1', 'n2', 'n3', 'n5', 'n6']:
f1.add_arc(state, 'n4', (letter), ('4'))
f1.add_arc('n4', 'n4', (letter), ())
elif letter in set('mn'):
for state in ['q2', 'n1', 'n2', 'n3', 'n4', 'n6']:
f1.add_arc(state, 'n5', (letter), ('5'))
f1.add_arc('n5', 'n5', (letter), ())
elif letter in set('r'):
for state in ['q2', 'n1', 'n2', 'n3', 'n4', 'n5']:
f1.add_arc(state, 'n6', (letter), ('6'))
f1.add_arc('n6', 'n6', (letter), ())
return f1
def letters_to_numbers():
"""
Returns an FST that converts letters to numbers as specified by
the soundex algorithm
"""
# Let's define our first FST
f1 = FST('soundex-generate')
# Indicate that '1' is the initial state
states = ['q1', 'q2', 'n1', 'n2', 'n3', 'n4', 'n5', 'n6']
for state in states:
f1.add_state(state)
f1.initial_state = 'q1'
# Set all the final states
for state in ['q2', 'n1', 'n2', 'n3', 'n4', 'n5', 'n6']:
f1.set_final(state)
# Add the rest of the arcs
for letter in string.ascii_lowercase:
f1.add_arc('q1', 'q2', (letter), (letter))
if letter in set('aehiouwy'):
for state in ['q2', 'n1', 'n2', 'n3', 'n4', 'n5', 'n6']:
f1.add_arc(state, state, (letter), ())
else:
if letter in set('bfpv'):
for state in ['q2', 'n2', 'n3', 'n4', 'n5', 'n6']:
f1.add_arc(state, 'n1', (letter), ('1'))
f1.add_arc('n1', 'n1', (letter), ())
elif letter in set('cgjkqsxz'):
for state in ['q2', 'n1', 'n3', 'n4', 'n5', 'n6']:
f1.add_arc(state, 'n2', (letter), ('2'))
f1.add_arc('n2', 'n2', (letter), ())
elif letter in set('dt'):
for state in ['q2', 'n1', 'n2', 'n4', 'n5', 'n6']:
f1.add_arc(state, 'n3', (letter), ('3'))
f1.add_arc('n3', 'n3', (letter), ())
elif letter in set('l'):
for state in ['q2', 'n1', 'n2', 'n3', 'n5', 'n6']:
f1.add_arc(state, 'n4', (letter), ('4'))
f1.add_arc('n4', 'n4', (letter), ())
elif letter in set('mn'):
for state in ['q2', 'n1', 'n2', 'n3', 'n4', 'n6']:
f1.add_arc(state, 'n5', (letter), ('5'))
f1.add_arc('n5', 'n5', (letter), ())
elif letter in set('r'):
for state in ['q2', 'n1', 'n2', 'n3', 'n4', 'n5']:
f1.add_arc(state, 'n6', (letter), ('6'))
f1.add_arc('n6', 'n6', (letter), ())
return f1
def generate(self, analysis):
"""Generate the morphologically correct word
e.g.
p = Parser()
analysis = ['p','a','n','i','c','+past form']
p.generate(analysis)
---> 'panicked'
"""
# Let's define our first FST
f1 = FST('morphology-generate')
# Indicate initial and final states
f1.add_state('start')
f1.add_state('vowel')
f1.add_state('consonant')
f1.add_state('c')
f1.add_state('form_1')
f1.add_state('form_2')
f1.initial_state = 'start'
f1.set_final('form_1')
f1.set_final('form_2')
# Generate
vowels = 'aeiou'
for vowel in vowels:
f1.add_arc('start', 'vowel', vowel, vowel)
f1.add_arc('vowel', 'vowel', vowel, vowel)
f1.add_arc('consonant', 'vowel', vowel, vowel)
f1.add_arc('c', 'vowel', vowel, vowel)
for letter in string.ascii_lowercase:
if letter in vowels:
continue
if not letter == 'c':
f1.add_arc('vowel', 'consonant', letter, letter)
f1.add_arc('start', 'consonant', letter, letter)
f1.add_arc('consonant', 'consonant', letter, letter)
f1.add_arc('c', 'consonant', letter, letter)
f1.add_arc('vowel', 'c', 'c', 'c')
f1.add_arc('c', 'form_1', '+past form', 'ked')
f1.add_arc('c', 'form_1', '+present participle form', 'king')
f1.add_arc('consonant', 'form_2', '+past form', 'ed')
f1.add_arc('consonant', 'form_2', '+present participle form', 'ing')
output = f1.transduce(analysis)[0]
return "".join(output)
def letters_to_numbers():
"""
Returns an FST that converts letters to numbers as specified by
the soundex algorithm
"""
# Let's define our first FST
f1 = FST('soundex-generate')
# Indicate that '1' is the initial state
f1.add_state('start')
f1.add_state('next')
f1.initial_state = 'start'
#create 0,1,2,3,4,5,6 states for all classes
for x in range(0,7):
f1.add_state(str(x))
f1.set_final(str(x))
list_0 = ['a','e','h','i','o','u','w','y']
list_1 = ['b','f','p','v']
list_2 = ['c','g','j','k','q','s','x','z']
list_3 = ['d','t']
list_4 = ['l']
list_5 = ['m','n']
list_6 = ['r']
all_lists = [list_0,list_1,list_2,list_3,list_4,list_5,list_6]
# Set all the final states
for index,item in enumerate(all_lists):
for letter in item:
f1.add_arc('start',str(index),(letter),(letter))
f1.add_arc('start',str(index),(letter.upper()),(letter.upper()))
for x in range(0,7):
for index,item in enumerate(all_lists):
for letter in item:
if x == index:
f1.add_arc(str(x),str(index),(letter),())
f1.add_arc(str(x),str(index),(letter.upper()),())
elif index == 0:
f1.add_arc(str(x),str(index),(letter),())
f1.add_arc(str(x),str(index),(letter.upper()),())
else:
f1.add_arc(str(x),str(index),(letter),(str(index)))
f1.add_arc(str(x),str(index),(letter.upper()),(str(index)))
return f1
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('initial')
f2.add_state('firstDigit')
f2.add_state('secondDigit')
f2.add_state('thirdDigit')
f2.initial_state = 'initial'
f2.set_final('initial')
f2.set_final('firstDigit')
f2.set_final('secondDigit')
f2.set_final('thirdDigit')
source = ['initial', 'firstDigit', 'secondDigit', 'thirdDigit']
destination = ['firstDigit', 'secondDigit', 'thirdDigit', 'thirdDigit']
# Add the arcs
for letter in string.letters:
f2.add_arc('initial', 'initial', (letter), (letter))
for cur, next in zip(source, destination):
f2 = addTruncateStates(cur, next, f2)
return f2
def add_zero_padding():
# Now, the third fst - the zero-padding fst
f3 = FST('soundex-padzero')
f3.add_state('start')
f3.initial_state = 'start'
for x in range(4):
f3.add_state(str(x))
f3.set_final(str(3))
# Add the arcs
f3.add_arc(str(0),str(1),(''),('0'))
f3.add_arc('start','1',(''),('0'))
f3.add_arc(str(1),str(2),(''),('0'))
f3.add_arc(str(2),str(3),(''),('0'))
for letter in string.letters:
f3.add_arc('start', '0', (letter), (letter))
for n in range(10):
f3.add_arc('start','1',(str(n)),(str(n)))
for x in range(3):
for n in range(10):
f3.add_arc(str(x), str(x+1), (str(n)), (str(n)))
for n in range(10):
f3.add_arc(str(3),str(3),(str(n)),())
'''
f3.add_state('1')
f3.add_state('1a')
f3.add_state('1b')
f3.add_state('2')
f3.initial_state = '1'
f3.set_final('2')
for letter in string.letters:
f3.add_arc('1', '1', (letter), (letter))
for number in xrange(10):
f3.add_arc('1', '1', (str(number)), (str(number)))
f3.add_arc('1', '1a', (), ('0'))
f3.add_arc('1a', '1b', (), ('0'))
f3.add_arc('1b', '2', (), ('0'))
'''
return f3
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('1')
f2.initial_state = '1'
f2.set_final('1')
return f2
def letters_to_numbers():
"""
Returns an FST that converts letters to numbers as specified by
the soundex algorithm
"""
# Let's define our first FST
f1 = FST('soundex-generate')
# Indicate that '1' is the initial state
f1.add_state('start')
f1.add_state('next')
f1.initial_state = 'start'
# Set all the final states
f1.set_final('next')
return f1
def letters_to_numbers():
"""
Returns an FST that converts letters to numbers as specified by
the soundex algorithm
"""
# Let's define our first FST
f1 = FST('soundex-generate')
# Indicate that '1' is the initial state
f1.add_state('start')
f1.add_state('next')
f1.initial_state = 'start'
# Set all the final states
f1.set_final('next')
# Add the rest of the arcs
for letter in string.ascii_lowercase:
f1.add_arc('start', 'next', (letter), (letter))
f1.add_arc('next', 'next', (letter), '0')
return f1
def truncate_to_three_digits():
"""
Create an FST that will truncate a soundex string to three digits
"""
# Ok so now let's do the second FST, the one that will truncate
# the number of digits to 3
f2 = FST('soundex-truncate')
# Indicate initial and final states
f2.add_state('1')
f2.initial_state = '1'
f2.set_final('1')
# Add the arcs
for letter in string.letters:
f2.add_arc('1', '1', (letter), (letter))
for n in range(10):
f2.add_arc('1', '1', str(n), str(n))
return f2
def french_count():
f = FST('french')
f.add_state('start')
f.initial_state = 'start'
f.add_state('1stzero')
f.add_state('tens')
f.add_state('seventeen')
f.add_state('final_seventeen')
f.add_state('eighteen')
f.add_state('final_eighteen')
f.add_state('nineteen')
f.add_state('final_nineteen')
f.add_state('zero')
f.add_state('ones')
f.add_state('20-69')
f.add_state('70-ten')
f.add_state('80s')
f.add_state('90s')
f.add_state('100s')
f.add_state('et')
f.add_state('10-et')
f.add_state('et-un')
f.add_state('et-onze')
f.set_final('zero')
f.set_final('ones')
f.set_final('tens')
f.set_final('final_seventeen')
f.set_final('final_eighteen')
f.set_final('final_nineteen')
f.set_final('20-69')
f.set_final('70-ten')
f.set_final('80s')
f.set_final('90s')
f.set_final('et-un')
f.set_final('et-onze')
# 100 - 999
f.add_arc('start', '1stzero', '1', [kFRENCH_TRANS[100]])
for i in range(2, 10):
f.add_arc('start', '100s', str(i), [kFRENCH_TRANS[i]])
f.add_arc('100s', '1stzero', (), [kFRENCH_TRANS[100]])
# 0 - 9
f.add_arc('start', '1stzero', '0', [])
f.add_arc('1stzero', 'ones', '0', [])
for ii in range(1, 10):
f.add_arc('ones', 'ones', str(ii), [kFRENCH_TRANS[ii]])
f.add_arc('ones', 'ones', '0', [])
# for i in range(10):
# f.add_arc('ten-6', 'ten-6', str(i), kFRENCH_TRANS[(i+10])
# 10 - 16
f.add_arc('1stzero', 'tens', '1', [])
f.add_arc('tens', 'tens', '0', [kFRENCH_TRANS[10]])
f.add_arc('tens', 'tens', '1', [kFRENCH_TRANS[11]])
f.add_arc('tens', 'tens', '2', [kFRENCH_TRANS[12]])
f.add_arc('tens', 'tens', '3', [kFRENCH_TRANS[13]])
f.add_arc('tens', 'tens', '4', [kFRENCH_TRANS[14]])
f.add_arc('tens', 'tens', '5', [kFRENCH_TRANS[15]])
f.add_arc('tens', 'tens', '6', [kFRENCH_TRANS[16]])
f.add_arc('tens', 'seventeen', '7', [kFRENCH_TRANS[10]])
f.add_arc('seventeen', 'final_seventeen', (), [kFRENCH_TRANS[7]])
f.add_arc('tens', 'eighteen', '8', [kFRENCH_TRANS[10]])
f.add_arc('eighteen', 'final_eighteen', (), [kFRENCH_TRANS[8]])
f.add_arc('tens', 'nineteen', '9', [kFRENCH_TRANS[10]])
f.add_arc('nineteen', 'final_nineteen', (), [kFRENCH_TRANS[9]])
# 20 - 69
f.add_arc('1stzero', '20-69', '2', [kFRENCH_TRANS[20]])
f.add_arc('1stzero', '20-69', '3', [kFRENCH_TRANS[30]])
f.add_arc('1stzero', '20-69', '4', [kFRENCH_TRANS[40]])
f.add_arc('1stzero', '20-69', '5', [kFRENCH_TRANS[50]])
f.add_arc('1stzero', '20-69', '6', [kFRENCH_TRANS[60]])
# special cases:
for i in range(2, 10):
f.add_arc('20-69', '20-69', str(i), [kFRENCH_TRANS[i]])
# handles 20, 30 ... 60
for i in range(20, 60, 10):
f.add_arc('20-69', '20-69', '0', [])
# handles 21, 31, ... 61
f.add_arc('20-69', 'et', '1', [kFRENCH_AND])
f.add_arc('et', 'et-un', (),[kFRENCH_TRANS[1]])
# 70 - 79
f.add_arc('1stzero', '70-ten', '7', [kFRENCH_TRANS[60]])
f.add_arc('70-ten', '70-ten', '0', [kFRENCH_TRANS[10]])
# handle 71 here
f.add_arc('70-ten', '10-et', '1', [kFRENCH_AND])
f.add_arc('10-et', 'et-onze', (),[kFRENCH_TRANS[11]])
f.add_arc('70-ten', '70-ten', '2', [kFRENCH_TRANS[12]])
f.add_arc('70-ten', '70-ten', '3', [kFRENCH_TRANS[13]])
f.add_arc('70-ten', '70-ten', '4', [kFRENCH_TRANS[14]])
f.add_arc('70-ten', '70-ten', '5', [kFRENCH_TRANS[15]])
f.add_arc('70-ten', '70-ten', '6', [kFRENCH_TRANS[16]])
f.add_arc('70-ten', 'seventeen', '7', [kFRENCH_TRANS[10]])
f.add_arc('seventeen', 'final_seventeen', (), [kFRENCH_TRANS[7]])
f.add_arc('70-ten', 'eighteen', '8', [kFRENCH_TRANS[10]])
f.add_arc('eighteen', 'final_eighteen', (), [kFRENCH_TRANS[8]])
f.add_arc('70-ten', 'nineteen', '9', [kFRENCH_TRANS[10]])
f.add_arc('nineteen', 'final_nineteen', (), [kFRENCH_TRANS[9]])
# 80 - 89
f.add_arc('1stzero', '80s', '8', [kFRENCH_TRANS[4]])
f.add_arc('80s', 'ones', (), [kFRENCH_TRANS[20]])
f.add_arc('80s', '80s', '0', [kFRENCH_TRANS[20]])
# 90 - 99
f.add_arc('1stzero', '90s', '9', [kFRENCH_TRANS[4]])
f.add_arc('90s', 'tens', (), [kFRENCH_TRANS[20]])
return f
def letters_to_numbers():
"""
Returns an FST that converts letters to numbers as specified by
the soundex algorithm
"""
# Let's define our first FST
f1 = FST('soundex-generate')
# Indicate that '1' is the initial state
f1.add_state('start')
f1.add_state('next')
f1.add_state('one')
f1.add_state('two')
f1.add_state('three')
f1.add_state('four')
f1.add_state('five')
f1.add_state('six')
f1.initial_state = 'start'
# Set all the final states
f1.set_final('next')
f1.set_final('one')
f1.set_final('two')
f1.set_final('three')
f1.set_final('four')
f1.set_final('five')
f1.set_final('six')
list_one = ['b', 'f', 'p', 'v']
list_two = ['c', 'g', 'j', 'k', 'q', 's', 'x', 'z']
list_three = ['d', 't']
list_four = ['l']
list_five = ['m', 'n']
list_six = ['r']
vowels = ['a', 'e', 'h', 'i', 'o', 'u', 'w', 'y']
# Add the rest of the arcs
# changed string.ascii_lowercase to string.letters
for letter in string.letters:
f1.add_arc('start', 'next', (letter), (letter))
for letter in string.letters:
if letter in list_one:
f1.add_arc('next', 'one', (letter), '1')
elif letter in list_two:
f1.add_arc('next', 'two', (letter), '2')
elif letter in list_three:
f1.add_arc('next', 'three', (letter), '3')
elif letter in list_four:
f1.add_arc('next', 'four', (letter), '4')
elif letter in list_five:
f1.add_arc('next', 'five', (letter), '5')
elif letter in list_six:
f1.add_arc('next', 'six', (letter), '6')
else:
f1.add_arc('next', 'next', (letter), ())
for letter in string.letters:
if letter in list_two:
f1.add_arc('one', 'two', (letter), '2')
elif letter in list_three:
f1.add_arc('one', 'three', (letter), '3')
elif letter in list_four:
f1.add_arc('one', 'four', (letter), '4')
elif letter in list_five:
f1.add_arc('one', 'five', (letter), '5')
elif letter in list_six:
f1.add_arc('one', 'six', (letter), '6')
else:
f1.add_arc('one', 'one', (letter), ())
for letter in string.letters:
if letter in list_one:
f1.add_arc('two', 'one', (letter), '1')
elif letter in list_three:
f1.add_arc('two', 'three', (letter), '3')
elif letter in list_four:
f1.add_arc('two', 'four', (letter), '4')
elif letter in list_five:
f1.add_arc('two', 'five', (letter), '5')
elif letter in list_six:
f1.add_arc('two', 'six', (letter), '6')
else:
f1.add_arc('two', 'two', (letter), ())
for letter in string.letters:
if letter in list_one:
f1.add_arc('three', 'one', (letter), '1')
elif letter in list_two:
f1.add_arc('three', 'two', (letter), '2')
elif letter in list_four:
f1.add_arc('three', 'four', (letter), '4')
elif letter in list_five:
f1.add_arc('three', 'five', (letter), '5')
elif letter in list_six:
f1.add_arc('three', 'six', (letter), '6')
else:
f1.add_arc('three', 'three', (letter), ())
for letter in string.letters:
if letter in list_one:
f1.add_arc('four', 'one', (letter), '1')
elif letter in list_two:
f1.add_arc('four', 'two', (letter), '2')
elif letter in list_three:
f1.add_arc('four', 'three', (letter), '3')
elif letter in list_five:
f1.add_arc('four', 'five', (letter), '5')
elif letter in list_six:
f1.add_arc('four', 'six', (letter), '6')
else:
f1.add_arc('four', 'four', (letter), ())
for letter in string.letters:
if letter in list_one:
f1.add_arc('five', 'one', (letter), '1')
elif letter in list_two:
f1.add_arc('five', 'two', (letter), '2')
elif letter in list_three:
f1.add_arc('five', 'three', (letter), '3')
elif letter in list_four:
f1.add_arc('five', 'four', (letter), '4')
elif letter in list_six:
f1.add_arc('five', 'six', (letter), '6')
else:
f1.add_arc('five', 'five', (letter), ())
for letter in string.letters:
if letter in list_one:
f1.add_arc('six', 'one', (letter), '1')
elif letter in list_two:
f1.add_arc('six', 'two', (letter), '2')
elif letter in list_three:
f1.add_arc('six', 'three', (letter), '3')
elif letter in list_four:
f1.add_arc('six', 'four', (letter), '4')
elif letter in list_five:
f1.add_arc('six', 'five', (letter), '5')
else:
f1.add_arc('six', 'six', (letter), ())
return f1
def french_count():
f = FST('french')
f.add_state('start')
# one number and two trailing unknowns
f.add_state('n**')
# exception from state n**
f.add_state('n**+')
# two numbers and one trailing unknown
f.add_state('nn*')
# zero and two uknown digits trailing and so on
f.add_state('0**')
f.add_state('00*')
f.add_state('00n')
f.add_state('0n*')
f.add_state('0n*+')
f.add_state('0nn')
f.add_state('n00')
f.add_state('nnn')
f.add_state('nnn*')
f.add_state('*et*')
# vegasimal counting for 7 in ((0/n)n*)
f.add_state('0n*Vega7+')
f.add_state('0n*Vega7')
f.add_state('0nnVega7')
# vegasimal counting for 8 in ((0/n)n*)
f.add_state('0n*Vega8')
f.add_state('0n*Vega8+')
f.add_state('0nnVega8')
# vegasimal counting for 9 in ((0/n)n*)
f.add_state('0n*Vega9')
f.add_state('0n*Vega9+')
f.add_state('0n*Vega9++')
f.add_state('0nnVega9')
# set final states
f.set_final('00n')
f.set_final('0nn')
f.set_final('nnn')
f.set_final('n00')
f.set_final('0nnVega7')
f.set_final('0nnVega8')
f.set_final('0nnVega9')
# initial state
f.initial_state = 'start'
# remove initial zeroes
f.add_arc('start', '0**', '0', ())
f.add_arc('0**', '00*', '0', ())
for ii in xrange(10):
#from '0n*Vega8' to '0nnVega8
if ii != 0:
f.add_arc('0n*Vega8+', '0nnVega8', str(ii), [kFRENCH_TRANS[ii]])
elif ii == 0:
f.add_arc('0n*Vega8+', '0nnVega8', str(ii), ())
#from '0n*Vega7' to '0nnVega7' 7-9
if ii == 0 or ii == 7 or ii ==8 or ii == 9:
f.add_arc('0n*Vega7', '0n*Vega7+', (), [kFRENCH_TRANS[10]])
f.add_arc('0n*Vega7+', '0n*Vega7+', str(ii), [kFRENCH_TRANS[ii]])
#
f.add_arc('0n*Vega9+', '0n*Vega9++', (), [kFRENCH_TRANS[10]])
f.add_arc('0n*Vega9++', '0nnVega9', str(ii), [kFRENCH_TRANS[ii]])
if ii == 0:
f.add_arc('0n*Vega7+', '0nnVega7', '0', ())
f.add_arc('0n*Vega9++', '0nnVega9', '0', ())
elif ii == 7 or ii == 8 or ii == 9:
f.add_arc('0n*Vega7+', '0nnVega7', str(ii), [kFRENCH_TRANS[ii]])
#from '0n*Vega' to '0nnVega' 2-6
if ii == 2 or ii == 3 or ii ==4 or ii == 5 or ii == 6:
f.add_arc('0n*Vega7', '0nnVega7', str(ii), [kFRENCH_TRANS[ii+10]])
f.add_arc('0n*Vega9+', '0nnVega9', str(ii), [kFRENCH_TRANS[ii+10]])
if ii == 1:
f.add_arc('0**','0n*', str(ii), [kFRENCH_TRANS[10]])
f.add_arc('n**','0n*', str(ii), [kFRENCH_TRANS[10]])
f.add_arc('0n*Vega7', '0n*Vega7+', str(ii), [kFRENCH_AND])
f.add_arc('0n*Vega7+', '0nnVega7', str(ii), [kFRENCH_TRANS[ii+10]])
f.add_arc('0n*Vega9+', '0nnVega9', str(ii), [kFRENCH_TRANS[ii+10]])
#from '00*' to '00n'
f.add_arc('00*', '00n', str(ii), [kFRENCH_TRANS[ii]])
#from '*n*' to '*nn' 2-9
if ii != 0 and ii !=9:
f.add_arc('0n*','0nn', str(ii+1), [kFRENCH_TRANS[ii+1]])
f.add_arc('0n*+','0nn', str(ii), [kFRENCH_TRANS[ii]])
#from 'start' to 'nnn' 200,300,...,900
if ii != 0 and ii !=1:
f.add_arc('start','n**+', str(ii), [kFRENCH_TRANS[ii]])
f.add_arc('n**+', 'n**', (), [kFRENCH_TRANS[100]])
#from 'n**' to 'n0*' 0
if ii == 0:
f.add_arc('n**', 'n00', '00', ())
if ii == 1:
f.add_arc('start', 'n**', '1', [kFRENCH_TRANS[100]])
#from '*n*' to '*et*' 1
f.add_arc('0n*','*et*', '1', [kFRENCH_AND])
#from '*et*' to '*nn' 1
f.add_arc('*et*','0nn', (), [kFRENCH_TRANS[1]])
#from '0**' to '*nn' 10-16
for ii in xrange(10,17):
f.add_arc('0**','0nn', str(ii), [kFRENCH_TRANS[ii]])
f.add_arc('n**','0nn', str(ii), [kFRENCH_TRANS[ii]])
#from '0**' to '*nn' 20-60
for ii in xrange(2,7):
f.add_arc('0**', '0nn', str(ii*10), [kFRENCH_TRANS[ii*10]])
f.add_arc('n**', '0nn', str(ii*10), [kFRENCH_TRANS[ii*10]])
#from '0**', to *n*
f.add_arc('0**','0n*', str(ii), [kFRENCH_TRANS[ii*10]])
#from 'n**' to '0n*'
f.add_arc('n**', '0n*+', str(ii), [kFRENCH_TRANS[ii*10]])
for ii in xrange(7,10):
if ii == 7:
f.add_arc('0**', '0n*Vega7', str(ii), [kFRENCH_TRANS[60]])
f.add_arc('n**', '0n*Vega7', str(ii), [kFRENCH_TRANS[60]])
elif ii == 8:
f.add_arc('0**', '0n*Vega8', str(ii), [kFRENCH_TRANS[4]])
f.add_arc('n**', '0n*Vega8', str(ii), [kFRENCH_TRANS[4]])
f.add_arc('0n*Vega8', '0n*Vega8+', (), [kFRENCH_TRANS[20]])
elif ii == 9:
f.add_arc('0**', '0n*Vega9', str(ii), [kFRENCH_TRANS[4]])
f.add_arc('n**', '0n*Vega9', str(ii), [kFRENCH_TRANS[4]])
f.add_arc('0n*Vega9', '0n*Vega9+', (), [kFRENCH_TRANS[20]])
f.add_arc('n**', '0n*+', '0', ())
return f
def french_count():
f = FST('french')
f.add_state('0')
f.add_state('1')
f.add_state('2')
f.add_state('3')
f.add_state('4')
f.add_state('5')
f.add_state('6')
f.add_state('7')
f.add_state('8')
f.add_state('9')
f.add_state('10')
f.add_state('11')
f.add_state('12')
f.add_state('13')
f.add_state('14')
f.add_state('15')
f.add_state('16')
f.add_state('17')
f.add_state('18')
f.add_state('19')
f.add_state('20')
f.add_state('21')
f.add_state('22')
f.add_state('23')
f.add_state('24')
f.add_state('25')
f.initial_state = '0'
f.set_final('1')
f.set_final('3')
f.set_final('6')
f.set_final('7')
f.set_final('8')
f.set_final('9')
f.set_final('11')
f.set_final('13')
f.set_final('14')
f.set_final('18')
f.set_final('20')
zero = [0]
one = [1]
two_to_six = [2,3,4,5,6]
one_to_six = [1,2,3,4,5,6]
seven = [7]
seven_eight_nine = [7,8,9]
eight = [8]
nine = [9]
singles_all = [1,2,3,4,5,6,7,8,9]
singles = [2,3,4,5,6,7,8,9]
tens = [20,30,40,50]
# Edge from initial to final, if preceding zero in input
for i in zero:
# f.add_arc('0','9', str(i), [kFRENCH_TRANS[i]])
f.add_arc('0','0', str(i), ())
f.add_arc('4','6', str(i), ())
f.add_arc('5','8', str(i), ())
f.add_arc('0','9', str(i), [kFRENCH_TRANS[i]])
f.add_arc('10','11', str(i), [kFRENCH_TRANS[i+10]])
f.add_arc('12','13', str(i), [kFRENCH_TRANS[20]])
f.add_arc('16','18', str(i), [kFRENCH_TRANS[20],kFRENCH_TRANS[10]])
f.add_arc('17','19', str(i), ())
f.add_arc('19','9', str(i), ())
for i in one:
f.add_arc('0','2', str(i), ())
f.add_arc('17','2', str(i), ())
f.add_arc('0','17', str(i), [kFRENCH_TRANS[100]])
f.add_arc('0','5', str(i), [kFRENCH_TRANS[i*10]])
f.add_arc('17','5', str(i), [kFRENCH_TRANS[i*10]])
f.add_arc('4','7', str(i), [kFRENCH_AND, kFRENCH_TRANS[i]])
f.add_arc('10','11', str(i), [kFRENCH_AND, kFRENCH_TRANS[i+10]])
f.add_arc('12','14', str(i), [kFRENCH_TRANS[20], kFRENCH_AND, kFRENCH_TRANS[i]])
f.add_arc('16','20', str(i), [kFRENCH_TRANS[20], kFRENCH_AND, kFRENCH_TRANS[i+10]])
for i in one_to_six:
f.add_arc('2','3', str(i), [kFRENCH_TRANS[i+10]])
for i in two_to_six:
f.add_arc('0','4', str(i), [kFRENCH_TRANS[i*10]])
f.add_arc('17','4', str(i), [kFRENCH_TRANS[i*10]])
f.add_arc('10','11', str(i), [kFRENCH_TRANS[i+10]])
f.add_arc('16','20', str(i), [kFRENCH_TRANS[20],kFRENCH_TRANS[i+10]])
for i in singles:
f.add_arc('4','7', str(i), [kFRENCH_TRANS[i]])
f.add_arc('0','17', str(i), [kFRENCH_TRANS[i],kFRENCH_TRANS[100]])
f.add_arc('12','14', str(i), [kFRENCH_TRANS[20], kFRENCH_TRANS[i]])
for i in singles_all:
f.add_arc('0','1', str(i), [kFRENCH_TRANS[i]])
f.add_arc('19','1', str(i), [kFRENCH_TRANS[i]])
for i in seven_eight_nine:
f.add_arc('5','8', str(i), [kFRENCH_TRANS[i]])
f.add_arc('10','11', str(i), [kFRENCH_TRANS[10], kFRENCH_TRANS[i]])
f.add_arc('16','20', str(i), [kFRENCH_TRANS[20], kFRENCH_TRANS[10], kFRENCH_TRANS[i]])
for i in seven:
f.add_arc('0','10',str(i), [kFRENCH_TRANS[60]])
f.add_arc('17','10',str(i), [kFRENCH_TRANS[60]])
for i in eight:
f.add_arc('0','12',str(i), [kFRENCH_TRANS[4]])
f.add_arc('17','12',str(i), [kFRENCH_TRANS[4]])
for i in nine:
f.add_arc('0','16',str(i), [kFRENCH_TRANS[4]])
f.add_arc('17','16',str(i), [kFRENCH_TRANS[4]])
return f
for o in output_list:
new_output = ''.join(o)
next_output_list.extend(fst.transduce(new_output))
output_list = next_output_list
return output_list
if __name__ == '__main__':
f1 = FST('test-generate')
# Indicate that '1' is the initial state
f1.add_state('start')
f1.add_state('next')
f1.initial_state = 'start'
# Set all the final states
f1.set_final('next')
# Add the rest of the arcs
for letter in ['A','B','C','D']:
f1.add_arc('start', 'next', letter, '1')
f1.add_arc('next', 'next', letter, '0')
f2 = FST('test-generate')
f2.add_state('start')
f2.add_state('next')
f2.initial_state = 'start'
f2.set_final('next')
f2.add_arc('start', 'next', '1', 'a')
f2.add_arc('start', 'next', '1', 'an')
f2.add_arc('next', 'next', '0', 'b')
def letters_to_numbers():
"""
Returns an FST that converts letters to numbers as specified by
the soundex algorithm
"""
# Let's define our first FST
f1 = FST('soundex-generate')
aeoy = ['a','e','h','i','o','u','w','y']
one = ['b','f','p','v']
two = ['c','g','j','k','q','s','x','z']
three = ['d','t']
four = ['l']
five = ['m','n']
six = ['r']
# Indicate that '1' is the initial state
f1.add_state('initial')
f1.add_state('0')
f1.add_state('1')
f1.add_state('2')
f1.add_state('3')
f1.add_state('4')
f1.add_state('5')
f1.add_state('6')
f1.initial_state = 'initial'
# Set all the final states
f1.set_final('0')
f1.set_final('1')
f1.set_final('2')
f1.set_final('3')
f1.set_final('4')
f1.set_final('5')
f1.set_final('6')
# Add the rest of the arcs
for letter in string.ascii_letters:
f1.add_arc('initial','0',(letter),(letter))
if letter in aeoy:
f1.add_arc('0','0', (letter), ())
f1.add_arc('1','0', (letter), ())
f1.add_arc('2','0', (letter), ())
f1.add_arc('3','0', (letter), ())
f1.add_arc('4','0', (letter), ())
f1.add_arc('5','0', (letter), ())
f1.add_arc('6','0', (letter), ())
else:
if letter in one:
f1.add_arc('0','1', (letter), '1')
f1.add_arc('2','1', (letter), '1')
f1.add_arc('3','1', (letter), '1')
f1.add_arc('4','1', (letter), '1')
f1.add_arc('5','1', (letter), '1')
f1.add_arc('6','1', (letter), '1')
f1.add_arc('1','0', (letter), ())
if letter in two:
f1.add_arc('0','2', (letter), '2')
f1.add_arc('1','2', (letter), '2')
f1.add_arc('3','2', (letter), '2')
f1.add_arc('4','2', (letter), '2')
f1.add_arc('5','2', (letter), '2')
f1.add_arc('6','2', (letter), '2')
f1.add_arc('2','0', (letter), ())
if letter in three:
f1.add_arc('0','3', (letter), '3')
f1.add_arc('1','3', (letter), '3')
f1.add_arc('2','3', (letter), '3')
f1.add_arc('4','3', (letter), '3')
f1.add_arc('5','3', (letter), '3')
f1.add_arc('6','3', (letter), '3')
f1.add_arc('3','0', (letter), ())
if letter in four:
f1.add_arc('0','4', (letter), '4')
f1.add_arc('1','4', (letter), '4')
f1.add_arc('2','4', (letter), '4')
f1.add_arc('3','4', (letter), '4')
f1.add_arc('5','4', (letter), '4')
f1.add_arc('6','4', (letter), '4')
f1.add_arc('4','0', (letter), ())
if letter in five:
f1.add_arc('0','5', (letter), '5')
f1.add_arc('1','5', (letter), '5')
f1.add_arc('2','5', (letter), '5')
f1.add_arc('3','5', (letter), '5')
f1.add_arc('4','5', (letter), '5')
f1.add_arc('6','5', (letter), '5')
f1.add_arc('5','0', (letter), ())
if letter in six:
f1.add_arc('0','6', (letter), '6')
f1.add_arc('1','6', (letter), '6')
f1.add_arc('2','6', (letter), '6')
f1.add_arc('3','6', (letter), '6')
f1.add_arc('4','6', (letter), '6')
f1.add_arc('5','6', (letter), '6')
f1.add_arc('6','0', (letter), ())
return f1
