mirror of
https://github.com/explosion/spaCy.git
synced 2024-11-11 12:18:04 +03:00
14 lines
426 B
Python
14 lines
426 B
Python
# coding: utf8
|
|
from __future__ import unicode_literals
|
|
|
|
import pytest
|
|
|
|
|
|
@pytest.mark.parametrize('text', ["Datum:2014-06-02\nDokument:76467"])
|
|
def test_issue886(en_tokenizer, text):
|
|
"""Test that token.idx matches the original text index for texts with newlines."""
|
|
doc = en_tokenizer(text)
|
|
for token in doc:
|
|
assert len(token.text) == len(token.text_with_ws)
|
|
assert text[token.idx] == token.text[0]
|