pyr0ball
8eef52a054
Implements Option A from the issue design: each cloud user gets their own
data directory (DATA_DIR/users/{user_id}/) with separate pagepiper.db,
pagepiper_vecs.db, uploads/, and books/. Local mode is unchanged.
Key changes:
- app/startup.py: extract apply_migrations, reembed_docs,
check_and_rebuild_vec_schema out of main.py (no circular imports)
- app/config.py: add LOCAL_USER_ID constant and user_data_dir() helper
- app/cloud_session.py: extract resolve_authenticated_user(); require_paid_tier
now returns user_id (str) instead of None
- app/deps.py: add UserCtx dataclass (db_path, vec_db_path, data_dir,
watch_dir, bm25) + get_user_ctx dependency; per-user startup guard runs
migrations + vec schema check once per process per user
- app/main.py: _bm25 singleton -> _bm25_map dict keyed by user_id;
add _get_bm25_for(); lifespan only runs startup checks in local mode
- app/api/library.py, search.py, chat.py: thread UserCtx through all
endpoints; remove module-level _mark_bm25_dirty injection pattern
- tests/conftest.py: override get_user_ctx in addition to get_db so all
endpoints get a consistent test UserCtx
65 lines
2.6 KiB
Python
65 lines
2.6 KiB
Python
# tests/test_search_api.py
|
|
"""Tests for POST /api/search — BM25 keyword search (MIT, no tier gate)."""
|
|
from __future__ import annotations
|
|
|
|
import sqlite3
|
|
|
|
|
|
def _add_chunks(db_path: str, doc_id: str, chunks: list[dict]) -> None:
|
|
conn = sqlite3.connect(db_path)
|
|
conn.execute(
|
|
"INSERT OR IGNORE INTO documents(id, title, file_path, status) VALUES (?,'Book','p.pdf','ready')",
|
|
[doc_id],
|
|
)
|
|
for c in chunks:
|
|
conn.execute(
|
|
"INSERT INTO page_chunks(doc_id, page_number, text, source, word_count) VALUES (?,?,?,?,?)",
|
|
[doc_id, c["page_number"], c["text"], "text_layer", len(c["text"].split())],
|
|
)
|
|
conn.commit()
|
|
conn.close()
|
|
|
|
|
|
def test_search_returns_results(client, test_db):
|
|
# BM25Okapi IDF is 0 when df == N/2 (e.g. 2 docs, 1 match → log(1.0) = 0).
|
|
# Add a 3rd unrelated chunk so relevant terms score above zero.
|
|
_add_chunks(test_db, "book-a", [
|
|
{"page_number": 1, "text": "Fireball deals 8d6 fire damage on a failed saving throw."},
|
|
{"page_number": 2, "text": "Cure Wounds restores hit points to a living creature."},
|
|
{"page_number": 3, "text": "Shield grants plus five to armor class until next turn."},
|
|
])
|
|
|
|
resp = client.post("/api/search", json={"query": "fireball fire damage"})
|
|
assert resp.status_code == 200
|
|
results = resp.json()
|
|
assert len(results) >= 1
|
|
assert results[0]["page_number"] == 1
|
|
assert results[0]["bm25_score"] > 0
|
|
assert "text_snippet" in results[0]
|
|
|
|
|
|
def test_search_empty_index_returns_empty(client):
|
|
resp = client.post("/api/search", json={"query": "anything"})
|
|
assert resp.status_code == 200
|
|
assert resp.json() == []
|
|
|
|
|
|
def test_search_filters_by_doc_ids(client, test_db):
|
|
# Three chunks so BM25Okapi IDF is non-zero for terms appearing in one doc.
|
|
_add_chunks(test_db, "book-a", [
|
|
{"page_number": 1, "text": "Grapple rules for melee attacks."},
|
|
{"page_number": 2, "text": "Shield spell protects from incoming blows."},
|
|
])
|
|
_add_chunks(test_db, "book-b", [{"page_number": 3, "text": "Grapple also applies to ranged attacks."}])
|
|
|
|
resp = client.post("/api/search", json={"query": "grapple", "doc_ids": ["book-a"]})
|
|
assert resp.status_code == 200
|
|
results = resp.json()
|
|
assert len(results) >= 1, "expected at least one grapple result from book-a"
|
|
assert all(r["doc_id"] == "book-a" for r in results)
|
|
|
|
|
|
def test_search_has_no_tier_gate(client):
|
|
# Search endpoint must return 200 with no PAGEPIPER_OLLAMA_URL set
|
|
resp = client.post("/api/search", json={"query": "anything"})
|
|
assert resp.status_code == 200 # Not 402
|