123 lines
4.6 KiB
Python
123 lines
4.6 KiB
Python
|
"""End-to-end тест: ingest → search → context → chat-stream.
|
|||
|
|
|
|||
|
|
Не мокает движок — ставит заглушку только на OpenCode-клиент.
|
|||
|
|
"""
|
|||
|
|
|
|||
|
|
import tempfile
|
|||
|
|
import unittest
|
|||
|
|
from pathlib import Path
|
|||
|
|
from unittest.mock import AsyncMock, MagicMock, patch
|
|||
|
|
|
|||
|
|
from src.rag.qmd import indexer as qmd_indexer
|
|||
|
|
from src.rag.qmd import query as qmd_query
|
|||
|
|
|
|||
|
|
|
|||
|
|
SAMPLE_MEETING = (
|
|||
|
|
"# Совещание 2026-06-10\n\n"
|
|||
|
|
"Участники: Иванов, Петров, Сидорова.\n\n"
|
|||
|
|
"## Повестка\n"
|
|||
|
|
"1. Ход строительства 3-й очереди.\n"
|
|||
|
|
"2. Авторизация подрядчиков в системе.\n"
|
|||
|
|
"3. Сроки сдачи.\n\n"
|
|||
|
|
"## Решения\n"
|
|||
|
|
"- Завершить фундамент до 15 июля.\n"
|
|||
|
|
"- Выдать JWT-токены подрядчикам.\n"
|
|||
|
|
"- Срок сдачи — 30 ноября 2027.\n"
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
|
|||
|
|
class E2ETestCase(unittest.IsolatedAsyncioTestCase):
|
|||
|
|
async def asyncSetUp(self):
|
|||
|
|
from src.rag.qmd.collections import get_project_collection_dir
|
|||
|
|
self._tmp = tempfile.TemporaryDirectory()
|
|||
|
|
self.tmp = Path(self._tmp.name)
|
|||
|
|
self.coll_dir = get_project_collection_dir("merakom", "2026", self.tmp)
|
|||
|
|
self.coll_dir.mkdir(parents=True, exist_ok=True)
|
|||
|
|
self.engines = []
|
|||
|
|
|
|||
|
|
async def asyncTearDown(self):
|
|||
|
|
from src.rag.engine import invalidate_engine
|
|||
|
|
for eng in self.engines:
|
|||
|
|
eng.close()
|
|||
|
|
invalidate_engine(self.coll_dir)
|
|||
|
|
import time
|
|||
|
|
time.sleep(0.05)
|
|||
|
|
try:
|
|||
|
|
self._tmp.cleanup()
|
|||
|
|
except (PermissionError, OSError):
|
|||
|
|
pass
|
|||
|
|
|
|||
|
|
async def test_e2e_meeting_index_search_chat(self):
|
|||
|
|
# 1. Ingest через Engine.index_file напрямую (в self.coll_dir).
|
|||
|
|
from src.rag.engine import get_or_create_engine
|
|||
|
|
eng = get_or_create_engine(self.coll_dir)
|
|||
|
|
eng.warmup()
|
|||
|
|
self.engines.append(eng)
|
|||
|
|
|
|||
|
|
body = self.tmp / "meeting.txt"
|
|||
|
|
summary = self.tmp / "meeting_summary.md"
|
|||
|
|
body.write_text(SAMPLE_MEETING, encoding="utf-8")
|
|||
|
|
summary.write_text("# Краткое\nОбсуждали строительство и JWT.", encoding="utf-8")
|
|||
|
|
r1 = eng.index_file(body)
|
|||
|
|
r2 = eng.index_file(summary)
|
|||
|
|
self.assertFalse(r1.skipped)
|
|||
|
|
self.assertFalse(r2.skipped)
|
|||
|
|
|
|||
|
|
# 2. Search через тот же engine.
|
|||
|
|
hits = eng.query("сдача объекта", limit=3, use_rerank=False)
|
|||
|
|
self.assertGreater(len(hits), 0)
|
|||
|
|
# top hit должен относиться к meeting.txt
|
|||
|
|
self.assertTrue(any("meeting.txt" in h.file_path for h in hits))
|
|||
|
|
|
|||
|
|
# 3. Chat-stream с подменой OpenCode.
|
|||
|
|
fake_chunks = [
|
|||
|
|
MagicMock(choices=[MagicMock(delta=MagicMock(content="Сдача "))]),
|
|||
|
|
MagicMock(choices=[MagicMock(delta=MagicMock(content="30 ноября "))]),
|
|||
|
|
MagicMock(choices=[MagicMock(delta=MagicMock(content="2027."))]),
|
|||
|
|
]
|
|||
|
|
|
|||
|
|
class _FakeStream:
|
|||
|
|
def __init__(self):
|
|||
|
|
self._i = 0
|
|||
|
|
|
|||
|
|
def __aiter__(self):
|
|||
|
|
return self
|
|||
|
|
|
|||
|
|
async def __anext__(self):
|
|||
|
|
if self._i >= len(fake_chunks):
|
|||
|
|
raise StopAsyncIteration
|
|||
|
|
item = fake_chunks[self._i]
|
|||
|
|
self._i += 1
|
|||
|
|
return item
|
|||
|
|
|
|||
|
|
async def _fake_create(*args, **kwargs):
|
|||
|
|
if kwargs.get("stream"):
|
|||
|
|
return _FakeStream()
|
|||
|
|
return MagicMock(choices=[MagicMock(message=MagicMock(content="final"))])
|
|||
|
|
|
|||
|
|
with patch("src.rag.qmd.query.AsyncOpenAI") as fake_cls:
|
|||
|
|
fake_instance = MagicMock()
|
|||
|
|
fake_instance.chat = MagicMock()
|
|||
|
|
fake_instance.chat.completions = MagicMock()
|
|||
|
|
fake_instance.chat.completions.create = AsyncMock(side_effect=_fake_create)
|
|||
|
|
fake_cls.return_value = fake_instance
|
|||
|
|
|
|||
|
|
events = []
|
|||
|
|
async for ev in qmd_query.qmd_chat_stream(
|
|||
|
|
question="Когда сдача?",
|
|||
|
|
org_slug="merakom",
|
|||
|
|
history=[],
|
|||
|
|
project_slug="2026",
|
|||
|
|
api_key="test-key",
|
|||
|
|
use_rerank=False,
|
|||
|
|
):
|
|||
|
|
events.append(ev)
|
|||
|
|
|
|||
|
|
types = [e["type"] for e in events]
|
|||
|
|
self.assertEqual(types[0], "context")
|
|||
|
|
self.assertIn("chunk", types)
|
|||
|
|
self.assertEqual(types[-1], "done")
|
|||
|
|
self.assertEqual(events[-1]["answer"], "Сдача 30 ноября 2027.")
|
|||
|
|
# Контекст непустой и содержит source-annotation
|
|||
|
|
self.assertIn("[source:", events[0]["context"])
|