turnstone/tests/context/test_embedder.py

53 lines
1.9 KiB
Python

"""Tests for app/context/embedder.py — graceful no-op without sqlite-vec."""
import sqlite3
from pathlib import Path
from unittest.mock import patch
import pytest
from app.context import embedder as emb_mod
@pytest.fixture
def db(tmp_path):
db_path = tmp_path / "t.db"
conn = sqlite3.connect(str(db_path))
conn.executescript("""
CREATE TABLE context_documents (
id TEXT PRIMARY KEY, filename TEXT NOT NULL, doc_type TEXT NOT NULL,
full_text TEXT NOT NULL, file_size INTEGER, uploaded_at TEXT NOT NULL
);
CREATE TABLE context_chunks (
id TEXT PRIMARY KEY, document_id TEXT NOT NULL
REFERENCES context_documents(id) ON DELETE CASCADE,
chunk_index INTEGER NOT NULL, text TEXT NOT NULL, embedding BLOB
);
INSERT INTO context_documents VALUES ('d1','test.md','markdown','hello',5,'2026-01-01T00:00:00+00:00');
INSERT INTO context_chunks VALUES ('c1','d1',0,'hello world',NULL);
""")
conn.commit()
conn.close()
return db_path
def test_embed_skipped_when_extension_absent(db):
with patch.object(emb_mod, "EMBEDDING_AVAILABLE", False):
count = emb_mod.embed_chunks(db, "d1", "http://localhost:11434")
assert count == 0
def test_embed_calls_ollama_when_available(db):
import httpx
class FakeResponse:
status_code = 200
def raise_for_status(self): pass
def json(self): return {"embedding": [0.1, 0.2, 0.3]}
with patch.object(emb_mod, "EMBEDDING_AVAILABLE", True), \
patch("app.context.embedder.httpx.post", return_value=FakeResponse()):
count = emb_mod.embed_chunks(db, "d1", "http://localhost:11434")
assert count == 1
# Verify blob was written
conn = sqlite3.connect(str(db))
row = conn.execute("SELECT embedding FROM context_chunks WHERE id='c1'").fetchone()
conn.close()
assert row[0] is not None