mirror of
https://github.com/explosion/spaCy.git
synced 2024-09-22 03:49:17 +03:00
17 lines
511 B
Python
17 lines
511 B
Python
|
# coding: utf-8
|
||
|
|
from __future__ import unicode_literals
|
||
|
|
|
||
|
|
import pytest
|
||
|
|
|
||
|
|
|
||
|
|
def test_eu_tokenizer_handles_long_text(eu_tokenizer):
|
||
|
|
text = """ta nere guitarra estrenatu ondoren"""
|
||
|
|
tokens = eu_tokenizer(text)
|
||
|
|
assert len(tokens) == 5
|
||
|
|
|
||
|
|
|
||
|
|
@pytest.mark.parametrize("text,length", [("milesker ederra joan zen hitzaldia plazer hutsa", 7), ("astelehen guztia sofan pasau biot", 5)])
|
||
|
|
def test_eu_tokenizer_handles_cnts(eu_tokenizer, text, length):
|
||
|
|
tokens = eu_tokenizer(text)
|
||
|
|
assert len(tokens) == length
|