spaCy/spacy/lang/hu/punctuation.py

41 lines
1.2 KiB
Python
Raw Normal View History

2017-03-12 15:07:28 +03:00
# coding: utf8
from __future__ import unicode_literals
2017-05-08 16:49:56 +03:00
_currency_symbols = r"\$ ¢ £ € ¥ ฿"
2017-01-14 17:51:59 +03:00
TOKENIZER_PREFIXES = (
[r'\+'] +
LIST_PUNCT +
LIST_ELLIPSES +
2017-05-08 16:49:56 +03:00
LIST_QUOTES)
2017-01-14 17:51:59 +03:00
TOKENIZER_SUFFIXES = (
LIST_PUNCT +
LIST_ELLIPSES +
LIST_QUOTES +
[
r'(?<=[0-9])\+',
r'(?<=°[FfCcKk])\.',
r'(?<=[0-9])(?:{c})'.format(c=CURRENCY),
r'(?<=[0-9])(?:{u})'.format(u=UNITS),
2017-05-08 16:49:56 +03:00
r'(?<=[{al}{p}{c}(?:{q})])\.'.format(al=ALPHA_LOWER, p=r'%²\-\)\]\+', q=QUOTES, c=_currency_symbols),
r'(?<=[{al})])-e'.format(al=ALPHA_LOWER)])
2017-01-14 17:51:59 +03:00
TOKENIZER_INFIXES = (
LIST_ELLIPSES +
[
r'(?<=[{al}])\.(?=[{au}])'.format(al=ALPHA_LOWER, au=ALPHA_UPPER),
r'(?<=[{a}]),(?=[{a}])'.format(a=ALPHA),
r'(?<=[{a}"])[:<>=](?=[{a}])'.format(a=ALPHA),
r'(?<=[{a}])--(?=[{a}])'.format(a=ALPHA),
r'(?<=[{a}]),(?=[{a}])'.format(a=ALPHA),
2017-05-08 16:49:56 +03:00
r'(?<=[{a}])([{q}\)\]\(\[])(?=[\-{a}])'.format(a=ALPHA, q=_QUOTES.replace("'", "").strip().replace(" ", ""))])
2017-05-08 23:29:04 +03:00
from ..char_classes import TOKENIZER_INFIXES, LIST_PUNCT LIST_ELLIPSES
from ..char_classes import LIST_QUOTES, CURRENCY, QUOTES, UNITS
from ..char_classes import ALPHA, ALPHA_LOWER, ALPHA_UPPER