def __init__(self, verbs):
self.identify_word_cache = {}
self.verb_cache = {}
self.verb_derivations, self.lemma2deriv_index = \
collect_verb_derivations(verbs)
self.deriv_index_picker = \
SuffixGeneralizingMap(self.lemma2deriv_index, min)
class Conjugator(object):
"""
Conjugates and un-conjugates verbs.
"""
def __init__(self, verbs):
self.identify_word_cache = {}
self.verb_cache = {}
self.verb_derivations, self.lemma2deriv_index = \
collect_verb_derivations(verbs)
self.deriv_index_picker = \
SuffixGeneralizingMap(self.lemma2deriv_index, min)
@staticmethod
def from_file(f):
vv = conjugations_from_file(f)
return Conjugator(vv)
def create_verb(self, lemma):
"""
lemma -> Verb
"""
verb = self.verb_cache.get(lemma)
if verb:
return verb
if lemma == MAGIC_INTS_LEMMA:
verb = Verb('0', '1', '2', map(str, range(3, 3 + 6)),
map(str, range(9, 9 + 6)))
else:
deriv_index = self.deriv_index_picker.get(lemma)
deriv = self.verb_derivations[deriv_index]
verb = deriv.derive_verb(lemma)
self.verb_cache[lemma] = verb
return verb
def identify_word(self, word, is_picky_about_verbs=True):
"""
conjugated word, is picky -> list of (lemma, verb field index)
"""
key = (word, is_picky_about_verbs)
lemmas_indexes = self.identify_word_cache.get(key)
if lemmas_indexes is not None:
return lemmas_indexes
# For each verb derivation, undo the conjugation in order to get the
# hypothetical original lemma.
#
# If the generalizing suffix map picks the same verb derivation we used
# for that lemma, it's a match.
lemmas_indexes = []
for i, deriv in enumerate(self.verb_derivations):
for lemma, field_index in deriv.identify_word(word):
if self.deriv_index_picker.get(lemma) == i:
lemmas_indexes.append((lemma, field_index))
# If we're picky, and we have some results that contain known verbs,
# forget about the results with unknown verbs.
if is_picky_about_verbs:
has_known = False
for lemma, field_index in lemmas_indexes:
if lemma in self.lemma2deriv_index:
has_known = True
break
if has_known:
lemmas_indexes = filter(
lambda (lemma, index): lemma in self.lemma2deriv_index,
lemmas_indexes)
lemmas_indexes.append((word, 0))
key = (word, is_picky_about_verbs)
self.identify_word_cache[key] = lemmas_indexes
return lemmas_indexes
