*rwet_midterm.py

# cribbed from Allison Parrish and Ross Goodwin
# https://gist.github.com/aparrish/ea3911c31cec8c858bd0/revisions
# https://gist.github.com/rossgoodwin/d45cba970add12c6190d

import random
from string import punctuation

# strip first ~50 lines of metadata
def not_with_semicolon(line):
    if not line.startswith(';;;') and not line[0] in punctuation:
        return True
    else:
        return False
# load and clean cmu dictionary
with open('cmudict-0.7b') as infile:
    lines = filter(not_with_semicolon, infile.read().strip().split('\n'))
#get rid of alternate pronunciations
lines = filter(lambda x: not '(' in x, lines)

#put the words in the dictionary
cmu_dict = dict()
for l in lines:
    word, phonemes = l.split('  ', 1)
    cmu_dict[word.lower()] = phonemes.split(' ')

#load markov'd file
# lines = [line.strip() for line in \
# 	open("/Users/user/Copy/itp/rwet/wk4/hw/markov-rhyme/drake-rhymes.txt").readlines()
# 	if len(line.strip()) > 0]
lines = [line.strip() for line in \
    open("/Users/user/Copy/itp/rwet/wk4/hw/markov-rhyme/drake-lyrics.txt").readlines()
    if len(line.strip()) > 0]

# get phones and lines for last words
last_words = list()
for line in lines:
    words = line.split()
    last_word = words[-1].strip('.,?!;:') # .strip() to remove any punctuation
    last_word_lower = last_word.lower()
    # if last word is in cmu dictionary (ie no slang, abbreviations, etc)
    if last_word_lower in cmu_dict:
        line_with_part = [cmu_dict[last_word_lower], line]
        last_words.append(line_with_part)
# print last_words

# generate alliterative lines
def get_alliteration_couplet(line_db):
    matching_phons = list()
    alliteration_lines = list()
    # grab random line to start
    first = random.choice(line_db)
    # create a list of lines that rhymes with the line randomly selected above
    for line in line_db:
        for p1, p2 in zip(first[0], line[0]):
            if p1 == p2:
                matching_phons.append(p1)
            else:
                break
            if len(matching_phons) > 1:
                alliteration_lines.append(first[1])
                alliteration_lines.append(line[1])
                break
    # print alliteration_lines
    second = random.choice(alliteration_lines)
    return [first[1], second]

# generate chorus
couplet1 = get_alliteration_couplet(last_words)
couplet2 = get_alliteration_couplet(last_words)
verse_line_one = couplet1[0]
verse_line_two = couplet1[1]
verse_line_three = couplet2[0]
verse_line_four = couplet2[1]
print verse_line_one
print verse_line_two
print verse_line_three
print verse_line_four

alliteration.py

# cribbed from Ross Goodwin
# https://gist.github.com/rossgoodwin/d45cba970add12c6190d

from string import punctuation

# strip first ~50 lines of metadata
def not_with_semicolon(line):
	if not line.startswith(';;;') and not line[0] in punctuation:
		return True
	else:
		return False

with open('cmudict-0.7b') as infile:
	lines = filter(not_with_semicolon, infile.read().strip().split('\n'))

#get rid of alternate pronunciations
lines = filter(lambda x: not '(' in x, lines)

cmu_dict = dict()

#put the words in the dictionary
for l in lines:
	word, phonemes = l.split('  ', 1)
	cmu_dict[word.lower()] = phonemes.split(' ')

def rhymes(word1, word2):
	#all phones in word
	try:
		phon1 = cmu_dict[word1]
		phon2 = cmu_dict[word2]
	except KeyError:
		raise Exception('Word not in dictionary.')

	matching_phons = list()
	# phon1[::1] puts the phones in reverse order
	# for p1, p2 in zip(phon1[::-1], phon2[::-1]):
	for p1, p2 in zip(phon1, phon2):
		#following line gets all matching phones, regarless of #/stress
		if p1 == p2:
			matching_phons.append(p1)
		else:
			break    

	if not matching_phons:
		return False    

	#if matching_phons[-1][-1] in set(['1', '2']):
	# change '> *'; 1-2 is best; higher gives better output
	if len(matching_phons) > 2:
		return True
	# can maybe do smt like else: if len(matching_phons) > 1: return True, and on down the line
	# until len(m_p) > 0, return true, else return false
	else:
		return False

def rhyme_finder(word):
	matches = list()
	for w in cmu_dict:
		if w != word and rhymes(w, word):
			matches.append(w)
	return matches

z = rhyme_finder('alligator')
print z
print len(z)

syllabic_rhyme.py

# cribbed from Ross Goodwin
# https://gist.github.com/rossgoodwin/d45cba970add12c6190d

from string import punctuation

# strip first ~50 lines of metadata
def not_with_semicolon(line):
	if not line.startswith(';;;') and not line[0] in punctuation:
		return True
	else:
		return False

with open('cmudict-0.7b') as infile:
	lines = filter(not_with_semicolon, infile.read().strip().split('\n'))

#get rid of alternate pronunciations
lines = filter(lambda x: not '(' in x, lines)

cmu_dict = dict()

#put the words in the dictionary
for l in lines:
	word, phonemes = l.split('  ', 1)
	cmu_dict[word.lower()] = phonemes.split(' ')

def rhymes(word1, word2):
	try:
		phon1 = cmu_dict[word1]
		phon2 = cmu_dict[word2]
	except KeyError:
		raise Exception('Word not in dictionary.')

	matching_phons = list()
	for p1, p2 in zip(phon1[::-1], phon2[::-1]):
		if p1 == p2:
			matching_phons.append(p1)
		else:
			break    
	if not matching_phons:
		return False    
	if matching_phons[-1][-1] in set(['0', '1', '2']):
		return True
	else:
		return False

def rhyme_finder(word):
	matches = list()
	for w in cmu_dict:
		if w != word and rhymes(w, word):
			matches.append(w)
	return matches

z = rhyme_finder('alligator')
print z
print len(z)

weak_rhyme.py

# cribbed from Ross Goodwin
# https://gist.github.com/rossgoodwin/d45cba970add12c6190d

from string import punctuation

# strip first ~50 lines of metadata
def not_with_semicolon(line):
	if not line.startswith(';;;') and not line[0] in punctuation:
		return True
	else:
		return False

with open('cmudict-0.7b') as infile:
	lines = filter(not_with_semicolon, infile.read().strip().split('\n'))

#get rid of alternate pronunciations
lines = filter(lambda x: not '(' in x, lines)

cmu_dict = dict()

#put the words in the dictionary
for l in lines:
	word, phonemes = l.split('  ', 1)
	cmu_dict[word.lower()] = phonemes.split(' ')

def rhymes(word1, word2):
	try:
		phon1 = cmu_dict[word1]
		phon2 = cmu_dict[word2]
	except KeyError:
		raise Exception('Word not in dictionary.')

	matching_phons = list()
	for p1, p2 in zip(phon1[::-1], phon2[::-1]):
		if p1 == p2:
			matching_phons.append(p1)
		else:
			break    
	if not matching_phons:
		return False    

	# matching_phons[-1][-1] is the stress number
	# matching_phons[-1] is the last matching phone
	if matching_phons[-1][-1] in set(['0']):
		#print '1: ', matching_phons
		#print '2: ', matching_phons[-1]
		#print '3: ', matching_phons[-1][-1]
		return True
	else:
		return False

def rhyme_finder(word):
	matches = list()
	for w in cmu_dict:
		if w != word and rhymes(w, word):
			matches.append(w)
	return matches

z = rhyme_finder('hammer')
print z
print len(z)

rhymes('abbeville', 'turbeville')