From 830c6710e074d06504fc02231b5d1a34fec61f21 Mon Sep 17 00:00:00 2001 From: Leon Luithlen Date: Wed, 13 Nov 2024 11:45:56 +0100 Subject: [PATCH] Fix chunk_by_word_test --- cognee/tests/unit/processing/chunks/chunk_by_word_test.py | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/cognee/tests/unit/processing/chunks/chunk_by_word_test.py b/cognee/tests/unit/processing/chunks/chunk_by_word_test.py index 54e19b162..38e5d9b5a 100644 --- a/cognee/tests/unit/processing/chunks/chunk_by_word_test.py +++ b/cognee/tests/unit/processing/chunks/chunk_by_word_test.py @@ -9,12 +9,17 @@ from cognee.tests.unit.processing.chunks.test_input import INPUT_TEXTS INPUT_TEXTS["python_code"], INPUT_TEXTS["chinese_text"] ]) - def test_chunk_by_word_isomorphism(input_text): chunks = chunk_by_word(input_text) reconstructed_text = "".join([chunk[0] for chunk in chunks]) assert reconstructed_text == input_text, f"texts are not identical: {len(input_text) = }, {len(reconstructed_text) = }" +@pytest.mark.parametrize("input_text", [ + INPUT_TEXTS["english_text"], + INPUT_TEXTS["english_lists"], + INPUT_TEXTS["python_code"], + INPUT_TEXTS["chinese_text"] +]) def test_chunk_by_word_splits(input_text): chunks = np.array(list(chunk_by_word(input_text))) space_test = np.array([" " not in chunk[0].strip() for chunk in chunks])