flexigo-support-bot / tests /rag /test_chunk.py
victor34593993's picture
deploy flexigo support bot
187966e verified
Raw
History Blame
747 Bytes
from __future__ import annotations
import pytest
from app.rag.chunk import chunk_text
def test_chunk_windows_and_overlap():
chunks = chunk_text("a" * 2500, size=1000, overlap=200)
assert all(len(c) <= 1000 for c in chunks)
assert len(chunks) == 4 # step=800 over 2500 chars
# consecutive chunks overlap by `overlap` chars
assert chunks[0][-200:] == chunks[1][:200]
def test_chunk_empty_returns_nothing():
assert chunk_text(" \n\t ") == []
assert chunk_text("") == []
def test_chunk_short_text_single_chunk():
assert chunk_text("hola mundo", size=1000, overlap=200) == ["hola mundo"]
def test_chunk_invalid_overlap():
with pytest.raises(ValueError):
chunk_text("abc", size=100, overlap=100)