def __init__(self, langName='', vowels='', consonants=''):
self.corpus = Corpus(langName, vowels, consonants)
if not vowels:
vowels = self.corpus.getVowels()
if not consonants:
consonants = self.corpus.getConsonants()
self.langName = langName
self.vowel = Vowel(vowels)
self.consonant = Consonant(consonants)
self.word = Word(self.vowel, self.consonant)
self.stats = {}
class Syntagmata(object):
'''
A python class for representing the statistical relationship of consonant
and vowel and consonant groups.
What operations do we want to support for this?
* obtaining statistical information from source text
* generating a dense representation of this in a data structure
* writing and reading this data structure
* the ability to combine mutliple source stats
o e.g., "I want this word to be 3 parts Russian-derived and 2 parts
Chinese-derived"
o this operation should be support via simply adding objects
together - addition will be the union of the two stat data
structures, where the frequencies (counts) are additive
o the percentage desired for a particular lanugage would be
indicated by integer multiplication, where the frequencies of the
data structure(s) would be multiplied by integers
* displaying the word parts
* displaying the stats for a given word part
>>> syn = Syntagmata(vowels='aeiouy', consonants='bcdfghjklmnpqrstvwxz')
>>> syn.vowel.letters
'aeiouy'
>>> syn.consonant.letters
'bcdfghjklmnpqrstvwxz'
>>> syn.isVowel('a'), syn.getLetterType('a').name
(True, 'VowelType')
>>> syn.isConsonant('b'), syn.getLetterType('b').name
(True, 'ConsonantType')
>>> syn.isWord('polychronopolous')
True
>>> syn.isWord('101 Main St.')
False
# check for pseudo-syllables
>>> syn.getPseudoSyllables('a')
['a']
>>> syn.getPseudoSyllables('the')
['the']
>>> syn.getPseudoSyllables('apple')
['appl', 'e']
>>> syn.getPseudoSyllables('frenetic')
['fre', 'ne', 'ti', 'c']
>>> syn.getPseudoSyllables('polychronopolous')
['po', 'ly', 'chro', 'no', 'po', 'lou', 's']
# test stats
>>> syn1 = Syntagmata(vowels='aeiouy', consonants='bcdfghjklmnpqrstvwxz')
>>> sourceText = "Polychronopolous is a delicious meal for the polyglot in all of us."
>>> stats = syn1.getStats(sourceText)
>>> finals = stats[FinalType.name].items()
>>> finals.sort()
>>> print finals
[('l', 1), ('r', 1), ('s', 2), ('t', 1)]
# some panagrams from http://www.p22.com/products/pangramcontest.html
>>> syn2 = Syntagmata(vowels='aeiouy', consonants='bcdfghjklmnpqrstvwxz')
>>> sourceText = """
... I jump quickly, shaving two dozen fox bare.
... The xylophone orchestra vowed to imbibe jugs of kumquat fizz.
... Knowledge of zymurgy and Bacchus justly pleases a quiet vixen.
... Zealous dominatrix whips frail, quivering boy with jockstrap!
... """
>>> stats = syn2.getStats(sourceText)
>>> finals = stats[FinalType.name].items()
>>> finals.sort()
>>> print finals
[('a', 1), ('ckly', 1), ('d', 1), ('dge', 1), ('e', 1), ('gs', 1), ('l', 1), ('mp', 1), ('n', 2), ('ne', 1), ('ng', 2), ('p', 1), ('ps', 1), ('re', 1), ('rgy', 1), ('s', 3), ('stly', 1), ('t', 2), ('th', 1), ('x', 2), ('zz', 1)]
# now test operations on the syntagmata
>>> syn3 = syn1 + syn1
>>> finals = syn3.stats[FinalType.name].items()
>>> finals.sort()
>>> print finals
[('l', 2), ('r', 2), ('s', 4), ('t', 2)]
>>> syn4 = syn1 * 1
>>> finals = syn4.stats[FinalType.name].items()
>>> finals.sort()
>>> print finals
[('l', 1), ('r', 1), ('s', 2), ('t', 1)]
>>> syn5 = syn1 * 2
>>> finals = syn5.stats[FinalType.name].items()
>>> finals.sort()
>>> print finals
[('l', 2), ('r', 2), ('s', 4), ('t', 2)]
>>> syn6 = syn1 * 3
>>> finals = syn6.stats[FinalType.name].items()
>>> finals.sort()
>>> print finals
[('l', 3), ('r', 3), ('s', 6), ('t', 3)]
>>> syn7 = syn1 * 10
>>> finals = syn7.stats[FinalType.name].items()
>>> finals.sort()
>>> print finals
[('l', 10), ('r', 10), ('s', 20), ('t', 10)]
>>> syn7 = syn1 * 10 + syn1 * 20 + syn1 * 3
>>> finals = syn7.stats[FinalType.name].items()
>>> finals.sort()
>>> print finals
[('l', 33), ('r', 33), ('s', 66), ('t', 33)]
'''
def __init__(self, langName='', vowels='', consonants=''):
self.corpus = Corpus(langName, vowels, consonants)
if not vowels:
vowels = self.corpus.getVowels()
if not consonants:
consonants = self.corpus.getConsonants()
self.langName = langName
self.vowel = Vowel(vowels)
self.consonant = Consonant(consonants)
self.word = Word(self.vowel, self.consonant)
self.stats = {}
def __add__(self, syntagObj):
vowels = self.vowel.letters + syntagObj.vowel.letters
consonants = self.consonant.letters + syntagObj.consonant.letters
newSyn = Syntagmata(vowels=vowels, consonants=consonants)
newSyn.stats = {
InitialType.name: Counter(),
MedialType.name: Counter(),
FinalType.name: Counter()}
itemSum = self.getStats().items() + syntagObj.getStats().items()
for key, data in itemSum:
for syllable, count in data.items():
try:
newSyn.stats[key][syllable] += count
except KeyError:
newSyn.stats[key].setdefault(syllable, count)
return newSyn
def __mul__(self, integer):
integer = int(integer)
if integer < 1:
msg = "You can only mulitiply by a positive integer greater than 0."
raise ValueError, msg
if integer == 1:
return self
newSyn = self + self
for i in xrange(integer - 2):
newSyn = newSyn + self
return newSyn
def isWord(self, word):
return self.word(word)
def isVowel(self, letter):
return self.vowel(letter)
def isConsonant(self, letter):
return self.consonant(letter)
def getLetterType(self, letter):
if self.isVowel(letter):
return VowelType
elif self.isConsonant(letter):
return ConsonantType
raise InvalidLetter
def getAlphabet(self):
return self.consonant.letters + self.vowel.letters
def makeWordList(self, sourceLines):
wordList = []
for line in sourceLines:
words = line.split()
for word in words:
word = word.strip().lower()
if self.isWord(word):
wordList.append(word.lower())
return wordList
def getPseudoSyllables(self, word):
part = ''
word = word.lower()
thisType = initialType = lastType = None
syllables = []
for letter in word:
try:
thisType = self.getLetterType(letter)
except InvalidLetter:
continue
if not part:
initialType = thisType
if thisType == initialType and thisType != lastType:
syllables.append(part)
part = ''
lastType = thisType
part += letter
# add last syllable
syllables.append(part)
# the first one in the list is always empty
return syllables[1:]
def generateStats(self, sourceText=''):
if not sourceText:
sourceText = self.corpus.getSource()
sourceLines = sourceText.split('\n')
initialData = Counter()
medialData = Counter()
finalData = Counter()
for word in self.makeWordList(sourceLines):
syllables = self.getPseudoSyllables(word)
initialData.updateCount(syllables[0])
final = ''
medials = []
if len(syllables) > 1:
finalData.updateCount(syllables[-1])
if len(syllables) > 2:
for medial in syllables[1:-1]:
medialData.updateCount(medial)
self.stats = {
InitialType.name: initialData,
MedialType.name: medialData,
FinalType.name: finalData}
return self.getStats()
def getStats(self, sourceText=''):
if self.stats:
return self.stats
try:
self.stats = self.corpus.getStats()
if self.stats:
return self.stats
except IOError:
pass
return self.generateStats(sourceText)
def makeWordPart(self, positionObj):
stats = self.getStats()
syllableData = stats[positionObj.name]
summedCounts = sum(syllableData.values())
choiceIndex = random.randint(1, summedCounts)
# next we need to get the count ranges for each pseudo-syllable so that
# we pick a syllable randomly but statistically accurately
ranges = []
lastEndpoint = 0
for syllable, count in syllableData.items():
thisRange = count + lastEndpoint
# if the randomly chosen index is between the last endpoint and the
# current endpoint, then we've got a match and will use the current
# syllable
if lastEndpoint < choiceIndex <= thisRange:
break
lastEndpoint += count
return syllable
def makeFinalPart(self):
return self.makeWordPart(FinalType)
def makeInitialPart(self):
return self.makeWordPart(InitialType)
def makeMedialPart(self):
return self.makeWordPart(MedialType)
def makeWord(self, syllableCount):
word = self.makeInitialPart()
for i in range(syllableCount - 2):
word += self.makeMedialPart()
word += self.makeFinalPart()
return word
def makeCVWord(self, syllableCount):
"""
This method forces a CV (consonant-vowel) pattern that is commonly
found in Japanese and Chinese phonology.
"""
word = ''
for i in range(int(syllableCount/2)):
word += self.makeInitialPart()
word += self.makeFinalPart()
return word
def writeStatsFile(self):
stats = self.getStats()
file = UTF8File(filename=self.corpus.statsFile, mode='w+')
file.write(str(stats))
file.close()
