andras/great-ai
1
0
Fork 0
Code Issues Pull requests 1 Projects Releases 4 Packages Wiki Activity Actions

Add files

Browse source 
Signed-off-by: András Schmelczer <andras@schmelczer.dev>
This commit is contained in:
Andras Schmelczer 2022-04-02 13:47:33 +02:00
commit 889e79174b
No known key found for this signature in database
GPG key ID: 39260B5B0614A13E
103 changed files with 24322 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

20
good_ai/tests/sus/test_lemmatize_token.py Normal file
View file

@ -0,0 +1,20 @@
import unittest
from src.sus.lemmatize_text import lemmatize_token
from src.sus.nlp import nlp
class TestLemmatizeToken(unittest.TestCase):
def test_simple(self) -> None:
token = nlp("Center")[0]
self.assertEqual(lemmatize_token(token), "centre")
self.assertEqual(lemmatize_token(token, add_negation=True), "centre")
self.assertEqual(lemmatize_token(token, add_part_of_speech=True), "centre_NOUN")
def test_punctuation(self) -> None:
token = nlp("This.")[1]
self.assertEqual(lemmatize_token(token), ".")
self.assertEqual(lemmatize_token(token, add_negation=True), ".")
self.assertEqual(lemmatize_token(token, add_part_of_speech=True), "._PUNCT")