def find_roots_for_partial_input(self, partial_input, whole_surface=None):
"""
@type partial_input: unicode
@type whole_surface: unicode
@rtype: list of Root
"""
assert partial_input and whole_surface
assert len(partial_input) <= len(whole_surface)
assert whole_surface.startswith(partial_input)
if len(whole_surface) == len(partial_input):
assert whole_surface == partial_input
root = partial_input
lemma = root
lemma_root = lemma
syntactic_category = SyntacticCategory.NOUN
secondary_syntactic_category = None
lexeme_attributes = set()
lexeme = DynamicLexeme(lemma, lemma_root, syntactic_category, secondary_syntactic_category,
lexeme_attributes)
phonetic_expectations = set()
phonetic_attributes = Phonetics.calculate_phonetic_attributes(partial_input, lexeme_attributes)
no_orthographics_root = DynamicRoot(root, lexeme, phonetic_expectations, phonetic_attributes)
if len(partial_input) < 2 <= len(whole_surface):
return []
if whole_surface == partial_input or len(partial_input) < 2:
return [no_orthographics_root]
last_vowel = Phonetics.get_last_vowel(partial_input)
if not last_vowel:
return [no_orthographics_root]
last_char = partial_input[-1]
first_char_after_partial_input = whole_surface[len(partial_input)]
if last_char.isupper() or first_char_after_partial_input.isupper():
return [no_orthographics_root]
roots = self._get_voicing_and_doubling_roots(partial_input, last_char, first_char_after_partial_input,
no_orthographics_root)
first_vowel_letter_after_partial_input = self._get_first_vowel(whole_surface[len(partial_input) - 1:])
if first_vowel_letter_after_partial_input:
if last_vowel.frontal != first_vowel_letter_after_partial_input.frontal:
for r in roots:
r.lexeme.attributes = set(r.lexeme.attributes)
r.lexeme.attributes.add(LexemeAttribute.InverseHarmony)
for r in roots:
phonetic_attributes = Phonetics.calculate_phonetic_attributes(r.str, r.lexeme.attributes)
r.phonetic_attributes = phonetic_attributes
return roots
def _get_possible_passive_roots(self, last_letter, partial_input, whole_surface, no_attr_root):
might_have_Passive_Il = (not last_letter.vowel and any([whole_surface.startswith(partial_input+s) for s in [u'il', u'ıl', u'ul', u'ül']])) or\
(last_letter.vowel and whole_surface.startswith(partial_input+ u'l'))
might_have_Passive_In = (not last_letter.vowel and any([whole_surface.startswith(partial_input+s) for s in [u'in', u'ın', u'un', u'ün']])) or\
(last_letter.vowel and whole_surface.startswith(partial_input+ u'n'))
might_have_Passive_InIl = (not last_letter.vowel and any([whole_surface.startswith(partial_input+s) for s in [u'inil', u'ınıl', u'unul', u'ünül']])) or\
(last_letter.vowel and any([whole_surface.startswith(partial_input+s) for s in [u'nil', u'nıl', u'nul', u'nül']]))
might_have_passives = {(LexemeAttribute.Passive_Il, might_have_Passive_Il),
(LexemeAttribute.Passive_In, might_have_Passive_In),
(LexemeAttribute.Passive_InIl, might_have_Passive_InIl)}
might_have_passives = filter(lambda t : t[1], might_have_passives)
passive_roots = set()
for passive_attr, might_have_happened in might_have_passives:
# cannot have other passives at the same time
# cannot have any other causative at the same time
# cannot have progressive vowel drop at the same time
# cannot have aorist_A or aorist_I at the same time
generated_root = no_attr_root._clone(True)
generated_root.lexeme.attributes = {passive_attr} if passive_attr else set()
generated_root.lexeme.phonetic_attributes = Phonetics.calculate_phonetic_attributes(partial_input, generated_root.lexeme.attributes)
passive_roots.add(generated_root)
return passive_roots
def get_phonetic_attributes(self):
if self.has_transitions():
suffix_so_far = self.get_surface_so_far()[len(self._root.str):]
if not suffix_so_far or suffix_so_far.isspace() or not suffix_so_far.isalnum():
return self._root.phonetic_attributes
else:
return Phonetics.calculate_phonetic_attributes(self.get_surface_so_far(), self.get_lexeme_attributes())
else:
return self._root.phonetic_attributes
def _set_lexeme_and_phonetic_attributes(self, generated_roots):
for r in generated_roots:
r.phonetic_attributes = Phonetics.calculate_phonetic_attributes(r.str, r.lexeme.attributes)
if r.str.endswith(u'd') and r.lexeme.root.endswith(u't'):
if LexemeAttribute.NoVoicing in r.lexeme.attributes:
r.lexeme.attributes.remove(LexemeAttribute.NoVoicing)
r.lexeme.attributes.add(LexemeAttribute.Voicing)
else:
if LexemeAttribute.Voicing in r.lexeme.attributes:
r.lexeme.attributes.remove(LexemeAttribute.Voicing)
r.lexeme.attributes.add(LexemeAttribute.NoVoicing)
def get_phonetic_attributes(self):
if self.has_transitions():
suffix_so_far = self.get_surface_so_far()[len(self._root.str):]
if not suffix_so_far or suffix_so_far.isspace(
) or not suffix_so_far.isalnum():
return self._root.phonetic_attributes
else:
return Phonetics.calculate_phonetic_attributes(
self.get_surface_so_far(), self.get_lexeme_attributes())
else:
return self._root.phonetic_attributes
def _set_lexeme_and_phonetic_attributes(self, generated_roots):
for r in generated_roots:
r.phonetic_attributes = Phonetics.calculate_phonetic_attributes(
r.str, r.lexeme.attributes)
if r.str.endswith(u'd') and r.lexeme.root.endswith(u't'):
if LexemeAttribute.NoVoicing in r.lexeme.attributes:
r.lexeme.attributes.remove(LexemeAttribute.NoVoicing)
r.lexeme.attributes.add(LexemeAttribute.Voicing)
else:
if LexemeAttribute.Voicing in r.lexeme.attributes:
r.lexeme.attributes.remove(LexemeAttribute.Voicing)
r.lexeme.attributes.add(LexemeAttribute.NoVoicing)
def _get_possible_causative_roots(self, partial_input, whole_surface,
no_attr_root):
# no voicing can happen on causative_t
might_have_Causative_t = whole_surface.startswith(partial_input + u't')
might_have_Causative_Ir = any([
whole_surface.startswith(partial_input + s)
for s in [u'ir', u'ır', u'ur', u'ür']
])
# no voicing can happen on causative_It
might_have_Causative_It = any([
whole_surface.startswith(partial_input + s)
for s in [u'it', u'ıt', u'ut', u'üt']
])
might_have_Causative_Ar = any([
whole_surface.startswith(partial_input + s)
for s in [u'ar', u'er']
])
might_have_Causative_dIr = any([whole_surface.startswith(partial_input+s) for s in [u'dir', u'dır', u'dur', u'dür']]) or\
any([whole_surface.startswith(partial_input+s) for s in [u'tir', u'tır', u'tur', u'tür']])
might_have_causatives = {
(LexemeAttribute.Causative_t, might_have_Causative_t),
(LexemeAttribute.Causative_Ir, might_have_Causative_Ir),
(LexemeAttribute.Causative_It, might_have_Causative_It),
(LexemeAttribute.Causative_Ar, might_have_Causative_Ar),
(LexemeAttribute.Causative_dIr, might_have_Causative_dIr)
}
might_have_causatives = filter(lambda t: t[1], might_have_causatives)
causative_roots = set()
for causative_attr, might_have_happened in might_have_causatives:
# cannot have other causatives at the same time
# cannot have any other passive at the same time
# cannot have progressive vowel drop at the same time
# cannot have aorist_A or aorist_I at the same time
generated_root = no_attr_root._clone(True)
generated_root.lexeme.attributes = {causative_attr
} if causative_attr else set()
generated_root.lexeme.phonetic_attributes = Phonetics.calculate_phonetic_attributes(
partial_input, generated_root.lexeme.attributes)
causative_roots.add(generated_root)
return causative_roots
def _get_possible_causative_roots(self, partial_input, whole_surface, no_attr_root):
# no voicing can happen on causative_t
might_have_Causative_t = whole_surface.startswith(partial_input + u't')
might_have_Causative_Ir = any([whole_surface.startswith(partial_input+s) for s in [u'ir', u'ır', u'ur', u'ür']])
# no voicing can happen on causative_It
might_have_Causative_It = any([whole_surface.startswith(partial_input+s) for s in [u'it', u'ıt', u'ut', u'üt']])
might_have_Causative_Ar = any([whole_surface.startswith(partial_input+s) for s in [u'ar', u'er']])
might_have_Causative_dIr = any([whole_surface.startswith(partial_input+s) for s in [u'dir', u'dır', u'dur', u'dür']]) or\
any([whole_surface.startswith(partial_input+s) for s in [u'tir', u'tır', u'tur', u'tür']])
might_have_causatives = {(LexemeAttribute.Causative_t, might_have_Causative_t),
(LexemeAttribute.Causative_Ir, might_have_Causative_Ir),
(LexemeAttribute.Causative_It, might_have_Causative_It),
(LexemeAttribute.Causative_Ar, might_have_Causative_Ar),
(LexemeAttribute.Causative_dIr, might_have_Causative_dIr)}
might_have_causatives = filter(lambda t : t[1], might_have_causatives)
causative_roots = set()
for causative_attr, might_have_happened in might_have_causatives:
# cannot have other causatives at the same time
# cannot have any other passive at the same time
# cannot have progressive vowel drop at the same time
# cannot have aorist_A or aorist_I at the same time
generated_root = no_attr_root._clone(True)
generated_root.lexeme.attributes = {causative_attr} if causative_attr else set()
generated_root.lexeme.phonetic_attributes = Phonetics.calculate_phonetic_attributes(partial_input, generated_root.lexeme.attributes)
causative_roots.add(generated_root)
return causative_roots
def _get_possible_passive_roots(self, last_letter, partial_input,
whole_surface, no_attr_root):
might_have_Passive_Il = (not last_letter.vowel and any([whole_surface.startswith(partial_input+s) for s in [u'il', u'ıl', u'ul', u'ül']])) or\
(last_letter.vowel and whole_surface.startswith(partial_input+ u'l'))
might_have_Passive_In = (not last_letter.vowel and any([whole_surface.startswith(partial_input+s) for s in [u'in', u'ın', u'un', u'ün']])) or\
(last_letter.vowel and whole_surface.startswith(partial_input+ u'n'))
might_have_Passive_InIl = (not last_letter.vowel and any([whole_surface.startswith(partial_input+s) for s in [u'inil', u'ınıl', u'unul', u'ünül']])) or\
(last_letter.vowel and any([whole_surface.startswith(partial_input+s) for s in [u'nil', u'nıl', u'nul', u'nül']]))
might_have_passives = {
(LexemeAttribute.Passive_Il, might_have_Passive_Il),
(LexemeAttribute.Passive_In, might_have_Passive_In),
(LexemeAttribute.Passive_InIl, might_have_Passive_InIl)
}
might_have_passives = filter(lambda t: t[1], might_have_passives)
passive_roots = set()
for passive_attr, might_have_happened in might_have_passives:
# cannot have other passives at the same time
# cannot have any other causative at the same time
# cannot have progressive vowel drop at the same time
# cannot have aorist_A or aorist_I at the same time
generated_root = no_attr_root._clone(True)
generated_root.lexeme.attributes = {passive_attr
} if passive_attr else set()
generated_root.lexeme.phonetic_attributes = Phonetics.calculate_phonetic_attributes(
partial_input, generated_root.lexeme.attributes)
passive_roots.add(generated_root)
return passive_roots
def ap(word, form_str, lexeme_attributes=None):
phonetic_attributes = Phonetics.calculate_phonetic_attributes(
word, lexeme_attributes)
word, application = Phonetics.apply(word, phonetic_attributes, form_str,
lexeme_attributes)
return word + application
def ap(word, form_str, lexeme_attributes=None):
phonetic_attributes = Phonetics.calculate_phonetic_attributes(word, lexeme_attributes)
word, application = Phonetics.apply(word, phonetic_attributes, form_str, lexeme_attributes)
return word + application
