mirror of
https://github.com/explosion/spaCy.git
synced 2025-11-04 09:57:26 +03:00
25 lines
812 B
Python
25 lines
812 B
Python
# coding: utf-8
|
|
from __future__ import unicode_literals
|
|
|
|
import pytest
|
|
from spacy.language import Language
|
|
from spacy.lang.punctuation import TOKENIZER_INFIXES
|
|
from spacy.lang.char_classes import ALPHA
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
"text,expected_tokens", [("l'avion", ["l'", "avion"]), ("j'ai", ["j'", "ai"])]
|
|
)
|
|
def test_issue768(text, expected_tokens):
|
|
"""Allow zero-width 'infix' token during the tokenization process."""
|
|
SPLIT_INFIX = r"(?<=[{a}]\')(?=[{a}])".format(a=ALPHA)
|
|
|
|
class FrenchTest(Language):
|
|
class Defaults(Language.Defaults):
|
|
infixes = TOKENIZER_INFIXES + [SPLIT_INFIX]
|
|
|
|
fr_tokenizer_w_infix = FrenchTest.Defaults.create_tokenizer()
|
|
tokens = fr_tokenizer_w_infix(text)
|
|
assert len(tokens) == 2
|
|
assert [t.text for t in tokens] == expected_tokens
|