| Server IP : 54.37.205.81 / Your IP : 216.73.216.76 Web Server : nginx/1.22.1 System : Linux vps-249481fa 6.1.0-50-cloud-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.176-1 (2026-07-02) x86_64 User : debian ( 1000) PHP Version : 8.2.32 Disable Function : NONE MySQL : OFF | cURL : ON | WGET : ON | Perl : ON | Python : OFF | Sudo : ON | Pkexec : OFF Directory : /usr/lib/python3/dist-packages/nltk/test/unit/ |
Upload File : |
import pytest
from nltk.util import everygrams
@pytest.fixture
def everygram_input():
"""Form test data for tests."""
return iter(["a", "b", "c"])
def test_everygrams_without_padding(everygram_input):
expected_output = [
("a",),
("a", "b"),
("a", "b", "c"),
("b",),
("b", "c"),
("c",),
]
output = list(everygrams(everygram_input))
assert output == expected_output
def test_everygrams_max_len(everygram_input):
expected_output = [
("a",),
("a", "b"),
("b",),
("b", "c"),
("c",),
]
output = list(everygrams(everygram_input, max_len=2))
assert output == expected_output
def test_everygrams_min_len(everygram_input):
expected_output = [
("a", "b"),
("a", "b", "c"),
("b", "c"),
]
output = list(everygrams(everygram_input, min_len=2))
assert output == expected_output
def test_everygrams_pad_right(everygram_input):
expected_output = [
("a",),
("a", "b"),
("a", "b", "c"),
("b",),
("b", "c"),
("b", "c", None),
("c",),
("c", None),
("c", None, None),
(None,),
(None, None),
(None,),
]
output = list(everygrams(everygram_input, max_len=3, pad_right=True))
assert output == expected_output
def test_everygrams_pad_left(everygram_input):
expected_output = [
(None,),
(None, None),
(None, None, "a"),
(None,),
(None, "a"),
(None, "a", "b"),
("a",),
("a", "b"),
("a", "b", "c"),
("b",),
("b", "c"),
("c",),
]
output = list(everygrams(everygram_input, max_len=3, pad_left=True))
assert output == expected_output