def __deserialize_item(self, stream: Stream) -> 'MorphBaseInfo':
from pullenti.ner.core.internal.SerializerHelper import SerializerHelper
ty = stream.readbyte()
res = (MorphBaseInfo() if ty == 0 else MorphWordForm())
res.class0_ = MorphClass._new53(
SerializerHelper.deserialize_short(stream))
res.case_ = MorphCase._new29(
SerializerHelper.deserialize_short(stream))
res.gender = Utils.valToEnum(
SerializerHelper.deserialize_short(stream), MorphGender)
res.number = Utils.valToEnum(
SerializerHelper.deserialize_short(stream), MorphNumber)
res.language = MorphLang._new56(
SerializerHelper.deserialize_short(stream))
if (ty == 0):
return res
wf = Utils.asObjectOrNull(res, MorphWordForm)
wf.normal_case = SerializerHelper.deserialize_string(stream)
wf.normal_full = SerializerHelper.deserialize_string(stream)
wf.undef_coef = SerializerHelper.deserialize_short(stream)
cou = SerializerHelper.deserialize_int(stream)
i = 0
while i < cou:
if (wf.misc is None):
wf.misc = MorphMiscInfo()
wf.misc.attrs.append(SerializerHelper.deserialize_string(stream))
i += 1
return res
def __init__(self, source: 'MorphCollection' = None) -> None:
super().__init__()
self.__m_class = MorphClass()
self.__m_gender = MorphGender.UNDEFINED
self.__m_number = MorphNumber.UNDEFINED
self.__m_case = MorphCase()
self.__m_language = MorphLang()
self.__m_voice = MorphVoice.UNDEFINED
self.__m_need_recalc = True
self.__m_items = None
if (source is None):
return
for it in source.items:
mi = None
if (isinstance(it, MorphWordForm)):
wf = MorphWordForm()
wf.copy_from_word_form(Utils.asObjectOrNull(it, MorphWordForm))
mi = (wf)
else:
mi = MorphBaseInfo()
mi.copy_from(it)
if (self.__m_items is None):
self.__m_items = list()
self.__m_items.append(mi)
self.__m_class = MorphClass._new53(source.__m_class.value)
self.__m_gender = source.__m_gender
self.__m_case = MorphCase._new29(source.__m_case.value)
self.__m_number = source.__m_number
self.__m_language = MorphLang._new56(source.__m_language.value)
self.__m_voice = source.__m_voice
self.__m_need_recalc = False
def __remove_items_morph_case(self, cas: 'MorphCase') -> None:
if (self.__m_items is None):
return
if (len(self.__m_items) == 0):
self.__m_case = ((self.__m_case) & cas)
for i in range(len(self.__m_items) - 1, -1, -1):
if (((self.__m_items[i].case_) & cas).is_undefined):
del self.__m_items[i]
self.__m_need_recalc = True
elif ((((self.__m_items[i].case_) & cas)) !=
self.__m_items[i].case_):
if (isinstance(self.__m_items[i], MorphWordForm)):
wf = MorphWordForm()
wf.copy_from_word_form(
Utils.asObjectOrNull(self.__m_items[i], MorphWordForm))
wf.case_ = (wf.case_) & cas
self.__m_items[i] = (wf)
else:
bi = MorphBaseInfo()
bi.copy_from(self.__m_items[i])
bi.case_ = (bi.case_) & cas
self.__m_items[i] = bi
self.__m_need_recalc = True
self.__m_need_recalc = True
def _createReferentToken(p : 'PersonReferent', begin : 'Token', end : 'Token', morph_ : 'MorphCollection', attrs : typing.List['PersonAttrToken'], ad : 'PersonAnalyzerData', for_attribute : bool, after_be_predicate : bool) -> 'ReferentToken':
from pullenti.ner.person.internal.PersonIdentityToken import PersonIdentityToken
if (p is None):
return None
has_prefix = False
if (attrs is not None):
for a in attrs:
if (a.typ == PersonAttrTerminType.BESTREGARDS):
has_prefix = True
else:
if (a.begin_char < begin.begin_char):
begin = a.begin_token
if (a.typ != PersonAttrTerminType.PREFIX):
if (a.age is not None):
p.addSlot(PersonReferent.ATTR_AGE, a.age, False, 0)
if (a.prop_ref is None):
p.addSlot(PersonReferent.ATTR_ATTR, a.value, False, 0)
else:
p.addSlot(PersonReferent.ATTR_ATTR, a, False, 0)
elif (a.gender == MorphGender.FEMINIE and not p.is_female):
p.is_female = True
elif (a.gender == MorphGender.MASCULINE and not p.is_male):
p.is_male = True
elif ((isinstance(begin.previous, TextToken)) and (begin.whitespaces_before_count < 3)):
if ((begin.previous).term == "ИП"):
a = PersonAttrToken(begin.previous, begin.previous)
a.prop_ref = PersonPropertyReferent()
a.prop_ref.name = "индивидуальный предприниматель"
p.addSlot(PersonReferent.ATTR_ATTR, a, False, 0)
begin = begin.previous
m0 = MorphCollection()
for it in morph_.items:
bi = MorphBaseInfo(it)
bi.number = MorphNumber.SINGULAR
if (bi.gender == MorphGender.UNDEFINED):
if (p.is_male and not p.is_female):
bi.gender = MorphGender.MASCULINE
if (not p.is_male and p.is_female):
bi.gender = MorphGender.FEMINIE
m0.addItem(bi)
morph_ = m0
if ((attrs is not None and len(attrs) > 0 and not attrs[0].morph.case_.is_undefined) and morph_.case_.is_undefined):
morph_.case_ = attrs[0].morph.case_
if (attrs[0].morph.number == MorphNumber.SINGULAR):
morph_.number = MorphNumber.SINGULAR
if (p.is_male and not p.is_female):
morph_.gender = MorphGender.MASCULINE
elif (p.is_female):
morph_.gender = MorphGender.FEMINIE
if (begin.previous is not None):
ttt = begin.previous
if (ttt.isValue("ИМЕНИ", "ІМЕНІ")):
for_attribute = True
else:
if (ttt.isChar('.') and ttt.previous is not None):
ttt = ttt.previous
if (ttt.whitespaces_after_count < 3):
if (ttt.isValue("ИМ", "ІМ")):
for_attribute = True
if (for_attribute):
return ReferentToken._new2329(p, begin, end, morph_, p._m_person_identity_typ)
if ((begin.previous is not None and begin.previous.is_comma_and and (isinstance(begin.previous.previous, ReferentToken))) and (isinstance(begin.previous.previous.getReferent(), PersonReferent))):
rt00 = Utils.asObjectOrNull(begin.previous.previous, ReferentToken)
ttt = rt00
while ttt is not None:
if (ttt.previous is None or not ((isinstance(ttt.previous.previous, ReferentToken)))):
break
if (not ttt.previous.is_comma_and or not ((isinstance(ttt.previous.previous.getReferent(), PersonReferent)))):
break
rt00 = (Utils.asObjectOrNull(ttt.previous.previous, ReferentToken))
ttt = (rt00)
if (isinstance(rt00.begin_token.getReferent(), PersonPropertyReferent)):
ok = False
if ((rt00.begin_token).end_token.next0_ is not None and (rt00.begin_token).end_token.next0_.isChar(':')):
ok = True
elif (rt00.begin_token.morph.number == MorphNumber.PLURAL):
ok = True
if (ok):
p.addSlot(PersonReferent.ATTR_ATTR, rt00.begin_token.getReferent(), False, 0)
if (ad is not None):
if (ad.overflow_level > 10):
return ReferentToken._new2329(p, begin, end, morph_, p._m_person_identity_typ)
ad.overflow_level += 1
attrs1 = None
has_position = False
open_br = False
t = end.next0_
first_pass3095 = True
while True:
if first_pass3095: first_pass3095 = False
else: t = t.next0_
if (not (t is not None)): break
if (t.is_table_control_char):
break
if (t.is_newline_before):
if (t.newlines_before_count > 2):
break
if (attrs1 is not None and len(attrs1) > 0):
break
ml = MailLine.parse(t, 0)
if (ml is not None and ml.typ == MailLine.Types.FROM):
break
if (t.chars.is_capital_upper):
attr1 = PersonAttrToken.tryAttach(t, (None if ad is None else ad.local_ontology), PersonAttrToken.PersonAttrAttachAttrs.NO)
ok1 = False
if (attr1 is not None):
if (has_prefix or attr1.is_newline_after or ((attr1.end_token.next0_ is not None and attr1.end_token.next0_.is_table_control_char))):
ok1 = True
else:
tt2 = t.next0_
while tt2 is not None and tt2.end_char <= attr1.end_char:
if (tt2.is_whitespace_before):
ok1 = True
tt2 = tt2.next0_
else:
ttt = PersonHelper.__correctTailAttributes(p, t)
if (ttt is not None and ttt != t):
t = ttt
end = t
continue
if (not ok1):
break
if (t.is_hiphen or t.isCharOf("_>|")):
continue
if (t.isValue("МОДЕЛЬ", None)):
break
tt = PersonHelper.__correctTailAttributes(p, t)
if (tt != t and tt is not None):
t = tt
end = t
continue
is_be = False
if (t.isChar('(') and t == end.next0_):
open_br = True
t = t.next0_
if (t is None):
break
pit1 = PersonItemToken.tryAttach(t, None, PersonItemToken.ParseAttr.NO, None)
if ((pit1 is not None and t.chars.is_capital_upper and pit1.end_token.next0_ is not None) and (isinstance(t, TextToken)) and pit1.end_token.next0_.isChar(')')):
if (pit1.lastname is not None):
inf = MorphBaseInfo._new2321(MorphCase.NOMINATIVE)
if (p.is_male):
inf.gender = Utils.valToEnum((inf.gender) | (MorphGender.MASCULINE), MorphGender)
if (p.is_female):
inf.gender = Utils.valToEnum((inf.gender) | (MorphGender.FEMINIE), MorphGender)
sur = PersonIdentityToken.createLastname(pit1, inf)
if (sur is not None):
p._addFioIdentity(sur, None, None)
t = pit1.end_token.next0_
end = t
continue
elif (t.is_comma):
t = t.next0_
if ((isinstance(t, TextToken)) and (t).isValue("WHO", None)):
continue
elif ((isinstance(t, TextToken)) and (t).is_verb_be):
t = t.next0_
elif (t.is_and and t.is_whitespace_after and not t.is_newline_after):
if (t == end.next0_):
break
t = t.next0_
elif (t.is_hiphen and t == end.next0_):
t = t.next0_
elif (t.isChar('.') and t == end.next0_ and has_prefix):
t = t.next0_
ttt2 = PersonHelper.createNickname(p, t)
if (ttt2 is not None):
end = ttt2
t = end
continue
if (t is None):
break
attr = None
attr = PersonAttrToken.tryAttach(t, (None if ad is None else ad.local_ontology), PersonAttrToken.PersonAttrAttachAttrs.NO)
if (attr is None):
if ((t is not None and t.getReferent() is not None and t.getReferent().type_name == "GEO") and attrs1 is not None and open_br):
continue
if ((t.chars.is_capital_upper and open_br and t.next0_ is not None) and t.next0_.isChar(')')):
if (p.findSlot(PersonReferent.ATTR_LASTNAME, None, True) is None):
p.addSlot(PersonReferent.ATTR_LASTNAME, t.getSourceText().upper(), False, 0)
t = t.next0_
end = t
if (t is not None and t.isValue("КОТОРЫЙ", None) and t.morph.number == MorphNumber.SINGULAR):
if (not p.is_female and t.morph.gender == MorphGender.FEMINIE):
p.is_female = True
p._correctData()
elif (not p.is_male and t.morph.gender == MorphGender.MASCULINE):
p.is_male = True
p._correctData()
break
if (attr.morph.number == MorphNumber.PLURAL):
break
if (attr.typ == PersonAttrTerminType.BESTREGARDS):
break
if (attr.is_doubt):
if (has_prefix):
pass
elif (t.is_newline_before and attr.is_newline_after):
pass
elif (t.previous is not None and ((t.previous.is_hiphen or t.previous.isChar(':')))):
pass
else:
break
if (not morph_.case_.is_undefined and not attr.morph.case_.is_undefined):
if (((morph_.case_) & attr.morph.case_).is_undefined and not is_be):
break
if (open_br):
if (PersonAnalyzer._tryAttachPerson(t, ad, False, 0, True) is not None):
break
if (attrs1 is None):
if (t.previous.is_comma and t.previous == end.next0_):
ttt = attr.end_token.next0_
if (ttt is not None):
if (ttt.morph.class0_.is_verb):
if (MiscHelper.canBeStartOfSentence(begin)):
pass
else:
break
attrs1 = list()
attrs1.append(attr)
if (attr.typ == PersonAttrTerminType.POSITION or attr.typ == PersonAttrTerminType.KING):
if (not is_be):
has_position = True
elif (attr.typ != PersonAttrTerminType.PREFIX):
if (attr.typ == PersonAttrTerminType.OTHER and attr.age is not None):
pass
else:
attrs1 = (None)
break
t = attr.end_token
if (attrs1 is not None and has_position and attrs is not None):
te1 = attrs[len(attrs) - 1].end_token.next0_
te2 = attrs1[0].begin_token
if (te1.whitespaces_after_count > te2.whitespaces_before_count and (te2.whitespaces_before_count < 2)):
pass
elif (attrs1[0].age is not None):
pass
elif (((te1.is_hiphen or te1.isChar(':'))) and not attrs1[0].is_newline_before and ((te2.previous.is_comma or te2.previous == end))):
pass
else:
for a in attrs:
if (a.typ == PersonAttrTerminType.POSITION):
te = attrs1[len(attrs1) - 1].end_token
if (te.next0_ is not None):
if (not te.next0_.isChar('.')):
attrs1 = (None)
break
if (attrs1 is not None and not has_prefix):
attr = attrs1[len(attrs1) - 1]
ok = False
if (attr.end_token.next0_ is not None and attr.end_token.next0_.chars.is_capital_upper):
ok = True
else:
rt = PersonAnalyzer._tryAttachPerson(attr.begin_token, ad, False, -1, False)
if (rt is not None and (isinstance(rt.referent, PersonReferent))):
ok = True
if (ok):
if (attr.begin_token.whitespaces_before_count > attr.end_token.whitespaces_after_count):
attrs1 = (None)
elif (attr.begin_token.whitespaces_before_count == attr.end_token.whitespaces_after_count):
rt1 = PersonAnalyzer._tryAttachPerson(attr.begin_token, ad, False, -1, False)
if (rt1 is not None):
attrs1 = (None)
if (attrs1 is not None):
for a in attrs1:
if (a.typ != PersonAttrTerminType.PREFIX):
if (a.age is not None):
p.addSlot(PersonReferent.ATTR_AGE, a.age, True, 0)
elif (a.prop_ref is None):
p.addSlot(PersonReferent.ATTR_ATTR, a.value, False, 0)
else:
p.addSlot(PersonReferent.ATTR_ATTR, a, False, 0)
end = a.end_token
if (a.gender != MorphGender.UNDEFINED and not p.is_female and not p.is_male):
if (a.gender == MorphGender.MASCULINE and not p.is_male):
p.is_male = True
p._correctData()
elif (a.gender == MorphGender.FEMINIE and not p.is_female):
p.is_female = True
p._correctData()
if (open_br):
if (end.next0_ is not None and end.next0_.isChar(')')):
end = end.next0_
crlf_cou = 0
t = end.next0_
first_pass3096 = True
while True:
if first_pass3096: first_pass3096 = False
else: t = t.next0_
if (not (t is not None)): break
if (t.is_table_control_char):
break
if (t.is_newline_before):
ml = MailLine.parse(t, 0)
if (ml is not None and ml.typ == MailLine.Types.FROM):
break
crlf_cou += 1
if (t.isCharOf(":,(") or t.is_hiphen):
continue
if (t.isChar('.') and t == end.next0_):
continue
r = t.getReferent()
if (r is not None):
if (r.type_name == "PHONE" or r.type_name == "URI" or r.type_name == "ADDRESS"):
ty = r.getStringValue("SCHEME")
if (r.type_name == "URI"):
if ((ty != "mailto" and ty != "skype" and ty != "ICQ") and ty != "http"):
break
p._addContact(r)
end = t
crlf_cou = 0
continue
if (isinstance(r, PersonIdentityReferent)):
p.addSlot(PersonReferent.ATTR_IDDOC, r, False, 0)
end = t
crlf_cou = 0
continue
if (r is not None and r.type_name == "ORGANIZATION"):
if (t.next0_ is not None and t.next0_.morph.class0_.is_verb):
break
if (begin.previous is not None and begin.previous.morph.class0_.is_verb):
break
if (t.whitespaces_after_count == 1):
break
exist = False
for s in p.slots:
if (s.type_name == PersonReferent.ATTR_ATTR and (isinstance(s.value, PersonPropertyReferent))):
pr = Utils.asObjectOrNull(s.value, PersonPropertyReferent)
if (pr.findSlot(PersonPropertyReferent.ATTR_REF, r, True) is not None):
exist = True
break
elif (s.type_name == PersonReferent.ATTR_ATTR and (isinstance(s.value, PersonAttrToken))):
pr = Utils.asObjectOrNull(s.value, PersonAttrToken)
if (pr.referent.findSlot(PersonPropertyReferent.ATTR_REF, r, True) is not None):
exist = True
break
if (not exist):
pat = PersonAttrToken(t, t)
pat.prop_ref = PersonPropertyReferent._new2291("сотрудник")
pat.prop_ref.addSlot(PersonPropertyReferent.ATTR_REF, r, False, 0)
p.addSlot(PersonReferent.ATTR_ATTR, pat, False, 0)
continue
if (r is not None):
break
if (not has_prefix or crlf_cou >= 2):
break
rt = t.kit.processReferent("PERSON", t)
if (rt is not None):
break
if (ad is not None):
ad.overflow_level -= 1
return ReferentToken._new2329(p, begin, end, morph_, p._m_person_identity_typ)
def try_attach_territory(
li: typing.List['TerrItemToken'],
ad: 'AnalyzerData',
attach_always: bool = False,
cits: typing.List['CityItemToken'] = None,
exists: typing.List['GeoReferent'] = None) -> 'ReferentToken':
if (li is None or len(li) == 0):
return None
ex_obj = None
new_name = None
adj_list = list()
noun = None
add_noun = None
rt = TerrAttachHelper.__try_attach_moscowao(li, ad)
if (rt is not None):
return rt
if (li[0].termin_item is not None
and li[0].termin_item.canonic_text == "ТЕРРИТОРИЯ"):
res2 = TerrAttachHelper.__try_attach_pure_terr(li, ad)
return res2
if (len(li) == 2):
if (li[0].rzd is not None and li[1].rzd_dir is not None):
rzd = GeoReferent()
rzd._add_name(li[1].rzd_dir)
rzd._add_typ_ter(li[0].kit.base_language)
rzd.add_slot(GeoReferent.ATTR_REF, li[0].rzd.referent, False,
0)
rzd.add_ext_referent(li[0].rzd)
return ReferentToken(rzd, li[0].begin_token, li[1].end_token)
if (li[1].rzd is not None and li[0].rzd_dir is not None):
rzd = GeoReferent()
rzd._add_name(li[0].rzd_dir)
rzd._add_typ_ter(li[0].kit.base_language)
rzd.add_slot(GeoReferent.ATTR_REF, li[1].rzd.referent, False,
0)
rzd.add_ext_referent(li[1].rzd)
return ReferentToken(rzd, li[0].begin_token, li[1].end_token)
can_be_city_before = False
adj_terr_before = False
if (cits is not None):
if (cits[0].typ == CityItemToken.ItemType.CITY):
can_be_city_before = True
elif (cits[0].typ == CityItemToken.ItemType.NOUN
and len(cits) > 1):
can_be_city_before = True
k = 0
k = 0
while k < len(li):
if (li[k].onto_item is not None):
if (ex_obj is not None or new_name is not None):
break
if (noun is not None):
if (k == 1):
if (noun.termin_item.canonic_text == "РАЙОН"
or noun.termin_item.canonic_text == "ОБЛАСТЬ"
or noun.termin_item.canonic_text == "СОЮЗ"):
if (isinstance(li[k].onto_item.referent,
GeoReferent)):
if (li[k].onto_item.referent.is_state):
break
ok = False
tt = li[k].end_token.next0_
if (tt is None):
ok = True
elif (tt.is_char_of(",.")):
ok = True
if (not ok):
ok = MiscLocationHelper.check_geo_object_before(
li[0].begin_token)
if (not ok):
adr = AddressItemToken.try_parse(
tt, None, False, False, None)
if (adr is not None):
if (adr.typ ==
AddressItemToken.ItemType.STREET):
ok = True
if (not ok):
break
if (li[k].onto_item is not None):
if (noun.begin_token.is_value("МО", None)
or noun.begin_token.is_value("ЛО", None)):
return None
ex_obj = li[k]
elif (li[k].termin_item is not None):
if (noun is not None):
break
if (li[k].termin_item.is_always_prefix and k > 0):
break
if (k > 0 and li[k].is_doubt):
if (li[k].begin_token == li[k].end_token
and li[k].begin_token.is_value("ЗАО", None)):
break
if (li[k].termin_item.is_adjective
or li[k].is_geo_in_dictionary):
adj_list.append(li[k])
else:
if (ex_obj is not None):
geo_ = Utils.asObjectOrNull(ex_obj.onto_item.referent,
GeoReferent)
if (geo_ is None):
break
if (ex_obj.is_adjective and
((li[k].termin_item.canonic_text == "СОЮЗ" or
li[k].termin_item.canonic_text == "ФЕДЕРАЦИЯ"))):
str0_ = str(ex_obj.onto_item)
if (not li[k].termin_item.canonic_text in str0_):
return None
if (li[k].termin_item.canonic_text == "РАЙОН"
or li[k].termin_item.canonic_text == "ОКРУГ"
or li[k].termin_item.canonic_text == "КРАЙ"):
tmp = io.StringIO()
for s in geo_.slots:
if (s.type_name == GeoReferent.ATTR_TYPE):
print("{0};".format(s.value),
end="",
file=tmp,
flush=True)
if (not li[k].termin_item.canonic_text
in Utils.toStringStringIO(tmp).upper()):
if (k != 1 or new_name is not None):
break
new_name = li[0]
new_name.is_adjective = True
new_name.onto_item = (None)
ex_obj = (None)
noun = li[k]
if (k == 0):
tt = TerrItemToken.try_parse(
li[k].begin_token.previous, None, True, False,
None)
if (tt is not None and tt.morph.class0_.is_adjective):
adj_terr_before = True
else:
if (ex_obj is not None):
break
if (new_name is not None):
break
new_name = li[k]
k += 1
name = None
alt_name = None
full_name = None
morph_ = None
if (ex_obj is not None):
if (ex_obj.is_adjective and not ex_obj.morph.language.is_en
and noun is None):
if (attach_always and ex_obj.end_token.next0_ is not None):
npt = NounPhraseHelper.try_parse(ex_obj.begin_token,
NounPhraseParseAttr.NO, 0,
None)
if (ex_obj.end_token.next0_.is_comma_and):
pass
elif (npt is None):
pass
else:
str0_ = StreetItemToken.try_parse(
ex_obj.end_token.next0_, None, False, None, False)
if (str0_ is not None):
if (str0_.typ == StreetItemType.NOUN
and str0_.end_token == npt.end_token):
return None
else:
cit = CityItemToken.try_parse(ex_obj.end_token.next0_,
None, False, None)
if (cit is not None
and ((cit.typ == CityItemToken.ItemType.NOUN
or cit.typ == CityItemToken.ItemType.CITY))):
npt = NounPhraseHelper.try_parse(
ex_obj.begin_token, NounPhraseParseAttr.NO, 0,
None)
if (npt is not None
and npt.end_token == cit.end_token):
pass
else:
return None
elif (ex_obj.begin_token.is_value("ПОДНЕБЕСНЫЙ", None)):
pass
else:
return None
if (noun is None and ex_obj.can_be_city):
cit0 = CityItemToken.try_parse_back(
ex_obj.begin_token.previous)
if (cit0 is not None
and cit0.typ != CityItemToken.ItemType.PROPERNAME):
return None
if (ex_obj.is_doubt and noun is None):
ok2 = False
if (TerrAttachHelper.__can_be_geo_after(
ex_obj.end_token.next0_)):
ok2 = True
elif (not ex_obj.can_be_surname and not ex_obj.can_be_city):
if ((ex_obj.end_token.next0_ is not None
and ex_obj.end_token.next0_.is_char(')')
and ex_obj.begin_token.previous is not None)
and ex_obj.begin_token.previous.is_char('(')):
ok2 = True
elif (ex_obj.chars.is_latin_letter
and ex_obj.begin_token.previous is not None):
if (ex_obj.begin_token.previous.is_value("IN", None)):
ok2 = True
elif (ex_obj.begin_token.previous.is_value(
"THE", None) and
ex_obj.begin_token.previous.previous is not None
and
ex_obj.begin_token.previous.previous.is_value(
"IN", None)):
ok2 = True
if (not ok2):
cit0 = CityItemToken.try_parse_back(
ex_obj.begin_token.previous)
if (cit0 is not None
and cit0.typ != CityItemToken.ItemType.PROPERNAME):
pass
elif (MiscLocationHelper.check_geo_object_before(
ex_obj.begin_token.previous)):
pass
else:
return None
name = ex_obj.onto_item.canonic_text
morph_ = ex_obj.morph
elif (new_name is not None):
if (noun is None):
return None
j = 1
while j < k:
if (li[j].is_newline_before and not li[0].is_newline_before):
if (BracketHelper.can_be_start_of_sequence(
li[j].begin_token, False, False)):
pass
else:
return None
j += 1
morph_ = noun.morph
if (new_name.is_adjective):
if (noun.termin_item.acronym == "АО"):
if (noun.begin_token != noun.end_token):
return None
if (new_name.morph.gender != MorphGender.FEMINIE):
return None
geo_before = None
tt0 = li[0].begin_token.previous
if (tt0 is not None and tt0.is_comma_and):
tt0 = tt0.previous
if (not li[0].is_newline_before and tt0 is not None):
geo_before = (Utils.asObjectOrNull(tt0.get_referent(),
GeoReferent))
if (Utils.indexOfList(li, noun, 0) < Utils.indexOfList(
li, new_name, 0)):
if (noun.termin_item.is_state):
return None
if (new_name.can_be_surname and geo_before is None):
if (((noun.morph.case_)
& new_name.morph.case_).is_undefined):
return None
if (MiscHelper.is_exists_in_dictionary(
new_name.begin_token, new_name.end_token,
(MorphClass.ADJECTIVE) | MorphClass.PRONOUN
| MorphClass.VERB)):
if (noun.begin_token != new_name.begin_token):
if (geo_before is None):
if (len(li) == 2 and
TerrAttachHelper.__can_be_geo_after(
li[1].end_token.next0_)):
pass
elif (len(li) == 3
and li[2].termin_item is not None
and TerrAttachHelper.__can_be_geo_after(
li[2].end_token.next0_)):
pass
elif (new_name.is_geo_in_dictionary):
pass
elif (new_name.end_token.is_newline_after):
pass
else:
return None
npt = NounPhraseHelper.try_parse(
new_name.end_token, NounPhraseParseAttr.PARSEPRONOUNS,
0, None)
if (npt is not None
and npt.end_token != new_name.end_token):
if (len(li) >= 3 and li[2].termin_item is not None
and npt.end_token == li[2].end_token):
add_noun = li[2]
else:
return None
rtp = new_name.kit.process_referent(
"PERSON", new_name.begin_token)
if (rtp is not None):
return None
name = ProperNameHelper.get_name_ex(
new_name.begin_token, new_name.end_token,
MorphClass.ADJECTIVE, MorphCase.UNDEFINED,
noun.termin_item.gender, False, False)
else:
ok = False
if (((k + 1) < len(li)) and li[k].termin_item is None
and li[k + 1].termin_item is not None):
ok = True
elif ((k < len(li)) and li[k].onto_item is not None):
ok = True
elif (k == len(li) and not new_name.is_adj_in_dictionary):
ok = True
elif (MiscLocationHelper.check_geo_object_before(
li[0].begin_token) or can_be_city_before):
ok = True
elif (MiscLocationHelper.check_geo_object_after(
li[k - 1].end_token, False)):
ok = True
elif (len(li) == 3 and k == 2):
cit = CityItemToken.try_parse(li[2].begin_token, None,
False, None)
if (cit is not None):
if (cit.typ == CityItemToken.ItemType.CITY
or cit.typ == CityItemToken.ItemType.NOUN):
ok = True
elif (len(li) == 2):
ok = TerrAttachHelper.__can_be_geo_after(
li[len(li) - 1].end_token.next0_)
if (not ok and not li[0].is_newline_before
and not li[0].chars.is_all_lower):
rt00 = li[0].kit.process_referent(
"PERSONPROPERTY", li[0].begin_token.previous)
if (rt00 is not None):
ok = True
if (noun.termin_item is not None
and noun.termin_item.is_strong
and new_name.is_adjective):
ok = True
if (noun.is_doubt and len(adj_list) == 0
and geo_before is None):
return None
name = ProperNameHelper.get_name_ex(
new_name.begin_token, new_name.end_token,
MorphClass.ADJECTIVE, MorphCase.UNDEFINED,
noun.termin_item.gender, False, False)
if (not ok and not attach_always):
if (MiscHelper.is_exists_in_dictionary(
new_name.begin_token, new_name.end_token,
(MorphClass.ADJECTIVE) | MorphClass.PRONOUN
| MorphClass.VERB)):
if (exists is not None):
for e0_ in exists:
if (e0_.find_slot(GeoReferent.ATTR_NAME,
name, True) is not None):
ok = True
break
if (not ok):
return None
full_name = "{0} {1}".format(
ProperNameHelper.get_name_ex(li[0].begin_token,
noun.begin_token.previous,
MorphClass.ADJECTIVE,
MorphCase.UNDEFINED,
noun.termin_item.gender,
False, False),
noun.termin_item.canonic_text)
else:
if (not attach_always or
((noun.termin_item is not None
and noun.termin_item.canonic_text == "ФЕДЕРАЦИЯ"))):
is_latin = noun.chars.is_latin_letter and new_name.chars.is_latin_letter
if (Utils.indexOfList(li, noun, 0) > Utils.indexOfList(
li, new_name, 0)):
if (not is_latin):
return None
if (not new_name.is_district_name
and not BracketHelper.can_be_start_of_sequence(
new_name.begin_token, False, False)):
if (len(adj_list) == 0
and MiscHelper.is_exists_in_dictionary(
new_name.begin_token, new_name.end_token,
(MorphClass.NOUN) | MorphClass.PRONOUN)):
if (len(li) == 2 and noun.is_city_region
and (noun.whitespaces_after_count < 2)):
pass
else:
return None
if (not is_latin):
if ((noun.termin_item.is_region
and not attach_always and
((not adj_terr_before or new_name.is_doubt)))
and not noun.is_city_region and
not noun.termin_item.is_specific_prefix):
if (not MiscLocationHelper.
check_geo_object_before(
noun.begin_token)):
if (not noun.is_doubt and noun.begin_token
!= noun.end_token):
pass
elif ((noun.termin_item.is_always_prefix
and len(li) == 2 and li[0] == noun)
and li[1] == new_name):
pass
else:
return None
if (noun.is_doubt and len(adj_list) == 0):
if (noun.termin_item.acronym == "МО"
or noun.termin_item.acronym == "ЛО"):
if (k == (len(li) - 1)
and li[k].termin_item is not None):
add_noun = li[k]
k += 1
elif (len(li) == 2 and noun == li[0]
and str(new_name).endswith("совет")):
pass
else:
return None
else:
return None
pers = new_name.kit.process_referent(
"PERSON", new_name.begin_token)
if (pers is not None):
return None
name = MiscHelper.get_text_value(new_name.begin_token,
new_name.end_token,
GetTextAttr.NO)
if (new_name.begin_token != new_name.end_token):
ttt = new_name.begin_token.next0_
while ttt is not None and ttt.end_char <= new_name.end_char:
if (ttt.chars.is_letter):
ty = TerrItemToken.try_parse(
ttt, None, False, False, None)
if ((ty is not None and ty.termin_item is not None
and noun is not None)
and ((noun.termin_item.canonic_text
in ty.termin_item.canonic_text
or ty.termin_item.canonic_text
in noun.termin_item.canonic_text))):
name = MiscHelper.get_text_value(
new_name.begin_token, ttt.previous,
GetTextAttr.NO)
break
ttt = ttt.next0_
if (len(adj_list) > 0):
npt = NounPhraseHelper.try_parse(adj_list[0].begin_token,
NounPhraseParseAttr.NO, 0,
None)
if (npt is not None and npt.end_token == noun.end_token):
alt_name = "{0} {1}".format(
npt.get_normal_case_text(None,
MorphNumber.UNDEFINED,
MorphGender.UNDEFINED,
False), name)
else:
if ((len(li) == 1 and noun is not None
and noun.end_token.next0_ is not None) and (isinstance(
noun.end_token.next0_.get_referent(), GeoReferent))):
g = Utils.asObjectOrNull(noun.end_token.next0_.get_referent(),
GeoReferent)
if (noun.termin_item is not None):
tyy = noun.termin_item.canonic_text.lower()
ooo = False
if (g.find_slot(GeoReferent.ATTR_TYPE, tyy, True)
is not None):
ooo = True
elif (tyy.endswith("район") and g.find_slot(
GeoReferent.ATTR_TYPE, "район", True) is not None):
ooo = True
if (ooo):
return ReferentToken._new734(g, noun.begin_token,
noun.end_token.next0_,
noun.begin_token.morph)
if ((len(li) == 1 and noun == li[0]
and li[0].termin_item is not None)
and TerrItemToken.try_parse(li[0].end_token.next0_, None,
True, False, None) is None and
TerrItemToken.try_parse(li[0].begin_token.previous, None,
True, False, None) is None):
if (li[0].morph.number == MorphNumber.PLURAL):
return None
cou = 0
str0_ = li[0].termin_item.canonic_text.lower()
tt = li[0].begin_token.previous
first_pass3158 = True
while True:
if first_pass3158: first_pass3158 = False
else: tt = tt.previous
if (not (tt is not None)): break
if (tt.is_newline_after):
cou += 10
else:
cou += 1
if (cou > 500):
break
g = Utils.asObjectOrNull(tt.get_referent(), GeoReferent)
if (g is None):
continue
ok = True
cou = 0
tt = li[0].end_token.next0_
first_pass3159 = True
while True:
if first_pass3159: first_pass3159 = False
else: tt = tt.next0_
if (not (tt is not None)): break
if (tt.is_newline_before):
cou += 10
else:
cou += 1
if (cou > 500):
break
tee = TerrItemToken.try_parse(tt, None, True, False,
None)
if (tee is None):
continue
ok = False
break
if (ok):
ii = 0
while g is not None and (ii < 3):
if (g.find_slot(GeoReferent.ATTR_TYPE, str0_, True)
is not None):
return ReferentToken._new734(
g, li[0].begin_token, li[0].end_token,
noun.begin_token.morph)
g = g.higher
ii += 1
break
return None
ter = None
if (ex_obj is not None and (isinstance(ex_obj.tag, GeoReferent))):
ter = (Utils.asObjectOrNull(ex_obj.tag, GeoReferent))
else:
ter = GeoReferent()
if (ex_obj is not None):
geo_ = Utils.asObjectOrNull(ex_obj.onto_item.referent,
GeoReferent)
if (geo_ is not None and not geo_.is_city):
ter._merge_slots2(geo_, li[0].kit.base_language)
else:
ter._add_name(name)
if (noun is None and ex_obj.can_be_city):
ter._add_typ_city(li[0].kit.base_language)
else:
pass
elif (new_name is not None):
ter._add_name(name)
if (alt_name is not None):
ter._add_name(alt_name)
if (noun is not None):
if (noun.termin_item.canonic_text == "АО"):
ter._add_typ(
("АВТОНОМНИЙ ОКРУГ" if li[0].kit.base_language.is_ua
else "АВТОНОМНЫЙ ОКРУГ"))
elif (noun.termin_item.canonic_text == "МУНИЦИПАЛЬНОЕ СОБРАНИЕ"
or noun.termin_item.canonic_text
== "МУНІЦИПАЛЬНЕ ЗБОРИ"):
ter._add_typ(("МУНІЦИПАЛЬНЕ УТВОРЕННЯ"
if li[0].kit.base_language.is_ua else
"МУНИЦИПАЛЬНОЕ ОБРАЗОВАНИЕ"))
elif (noun.termin_item.acronym == "МО"
and add_noun is not None):
ter._add_typ(add_noun.termin_item.canonic_text)
else:
if (noun.termin_item.canonic_text == "СОЮЗ"
and ex_obj is not None
and ex_obj.end_char > noun.end_char):
return ReferentToken._new734(ter, ex_obj.begin_token,
ex_obj.end_token,
ex_obj.morph)
ter._add_typ(noun.termin_item.canonic_text)
if (noun.termin_item.is_region and ter.is_state):
ter._add_typ_reg(li[0].kit.base_language)
if (ter.is_state and ter.is_region):
for a in adj_list:
if (a.termin_item.is_region):
ter._add_typ_reg(li[0].kit.base_language)
break
if (ter.is_state):
if (full_name is not None):
ter._add_name(full_name)
res = ReferentToken(ter, li[0].begin_token, li[k - 1].end_token)
if (noun is not None and noun.morph.class0_.is_noun):
res.morph = noun.morph
else:
res.morph = MorphCollection()
ii = 0
while ii < k:
for v in li[ii].morph.items:
bi = MorphBaseInfo()
bi.copy_from(v)
if (noun is not None):
if (bi.class0_.is_adjective):
bi.class0_ = MorphClass.NOUN
res.morph.add_item(bi)
ii += 1
if (li[0].termin_item is not None
and li[0].termin_item.is_specific_prefix):
res.begin_token = li[0].end_token.next0_
if (add_noun is not None and add_noun.end_char > res.end_char):
res.end_token = add_noun.end_token
if ((isinstance(res.begin_token.previous, TextToken))
and (res.whitespaces_before_count < 2)):
tt = Utils.asObjectOrNull(res.begin_token.previous, TextToken)
if (tt.term == "АР"):
for ty in ter.typs:
if ("республика" in ty or "республіка" in ty):
res.begin_token = tt
break
return res
def __try_parse_en(first: 'Token', typ: 'NounPhraseParseAttr',
max_char_pos: int) -> 'NounPhraseToken':
if (first is None):
return None
items = None
has_article = False
has_prop = False
has_misc = False
if (first.previous is not None
and first.previous.morph.class0_.is_preposition
and (first.whitespaces_before_count < 3)):
has_prop = True
t = first
first_pass3048 = True
while True:
if first_pass3048: first_pass3048 = False
else: t = t.next0_
if (not (t is not None)): break
if (max_char_pos > 0 and t.begin_char > max_char_pos):
break
if (not t.chars.is_latin_letter):
break
if (t != first and t.whitespaces_before_count > 2):
if ((((typ) & (NounPhraseParseAttr.MULTILINES))) !=
(NounPhraseParseAttr.NO)):
pass
elif (MiscHelper.is_eng_article(t.previous)):
pass
else:
break
tt = Utils.asObjectOrNull(t, TextToken)
if (t == first and tt is not None):
if (MiscHelper.is_eng_article(tt)):
has_article = True
continue
if (isinstance(t, ReferentToken)):
if ((((typ) & (NounPhraseParseAttr.REFERENTCANBENOUN))) == (
NounPhraseParseAttr.NO)):
break
elif (tt is None):
break
if ((t.is_value("SO", None) and t.next0_ is not None
and t.next0_.is_hiphen) and t.next0_.next0_ is not None):
if (t.next0_.next0_.is_value("CALL", None)):
t = t.next0_.next0_
continue
mc = t.get_morph_class_in_dictionary()
if (mc.is_conjunction or mc.is_preposition):
break
if (mc.is_pronoun or mc.is_personal_pronoun):
if ((((typ) & (NounPhraseParseAttr.PARSEPRONOUNS))) == (
NounPhraseParseAttr.NO)):
break
elif (mc.is_misc):
if (t.is_value("THIS", None) or t.is_value("THAT", None)):
has_misc = True
if ((((typ) & (NounPhraseParseAttr.PARSEPRONOUNS))) == (
NounPhraseParseAttr.NO)):
break
is_adj = False
if (((has_article or has_prop or has_misc)) and items is None):
pass
elif (isinstance(t, ReferentToken)):
pass
else:
if (not mc.is_noun and not mc.is_adjective):
if (mc.is_undefined and has_article):
pass
elif (items is None and mc.is_undefined
and t.chars.is_capital_upper):
pass
elif (mc.is_pronoun):
pass
elif (tt.term.endswith("EAN")):
is_adj = True
elif (MiscHelper.is_eng_adj_suffix(tt.next0_)):
pass
else:
break
if (mc.is_verb):
if (t.next0_ is not None and t.next0_.morph.class0_.is_verb
and (t.whitespaces_after_count < 2)):
pass
elif (t.chars.is_capital_upper
and not MiscHelper.can_be_start_of_sentence(t)):
pass
elif ((t.chars.is_capital_upper and mc.is_noun and
(isinstance(t.next0_, TextToken)))
and t.next0_.chars.is_capital_upper):
pass
elif (isinstance(t, ReferentToken)):
pass
else:
break
if (items is None):
items = list()
it = NounPhraseItem(t, t)
if (mc.is_noun):
it.can_be_noun = True
if (mc.is_adjective or mc.is_pronoun or is_adj):
it.can_be_adj = True
items.append(it)
t = it.end_token
if (len(items) == 1):
if (MiscHelper.is_eng_adj_suffix(t.next0_)):
mc.is_noun = False
mc.is_adjective = True
t = t.next0_.next0_
if (items is None):
return None
noun = items[len(items) - 1]
res = NounPhraseToken(first, noun.end_token)
res.noun = (noun)
res.morph = MorphCollection()
for v in noun.end_token.morph.items:
if (v.class0_.is_verb):
continue
if (v.class0_.is_proper and noun.begin_token.chars.is_all_lower):
continue
if (isinstance(v, MorphWordForm)):
wf = MorphWordForm()
wf.copy_from_word_form(Utils.asObjectOrNull(v, MorphWordForm))
if (has_article and v.number != MorphNumber.SINGULAR):
wf.number = MorphNumber.SINGULAR
res.morph.add_item(wf)
else:
bi = MorphBaseInfo()
bi.copy_from(v)
if (has_article and v.number != MorphNumber.SINGULAR):
bi.number = MorphNumber.SINGULAR
res.morph.add_item(bi)
if (res.morph.items_count == 0 and has_article):
res.morph.add_item(
MorphBaseInfo._new192(MorphClass.NOUN, MorphNumber.SINGULAR))
i = 0
while i < (len(items) - 1):
res.adjectives.append(items[i])
i += 1
return res