feat(ebook): migrate to async DB/HTTP and parallelize phrase pipeline

Convert the ebook-search web app to async end to end and add concurrency
to the protected-phrase extraction and judging pipeline so large books no
longer block the event loop or the UI.

ORM / infra:
- Add get_async_postgres_engine and factor shared URL/connect_args building
  into build_postgres_url (reused by the sync and async engine builders)
- Add async FastAPI session helpers (get_async_db, AsyncDbSession) with
  expire_on_commit=False to avoid implicit IO under asyncio

App:
- Use AsyncEngine/AsyncSession throughout routes, search, ingest, embeddings,
  answer, rerank and LLM calls; convert handlers to async
- Share a single httpx.AsyncClient in app state for LLM requests; size the
  connection pool for concurrent phrase-judging workers
- Add judge_tasks: run per-book judging as tracked background tasks so a
  book already being judged isn't double-queued

Protected phrases:
- Add a process pool (pool.py) and worker-count config
  (extraction/judge book/phrase workers) to parallelize candidate generation
  and judging
- Split admin actions into all/missing variants for generation and judging

Config:
- Add protected_phrase_extraction_workers, phrase_judge_book_workers,
  phrase_judge_phrase_workers
This commit is contained in:
2026-07-12 13:34:04 -04:00
parent 903ad749b0
commit 76a9c70a47
29 changed files with 1824 additions and 769 deletions
+15 -13
View File
@@ -11,15 +11,17 @@ from fastapi.responses import JSONResponse
from sqlalchemy import literal, select
from sqlalchemy.exc import SQLAlchemyError
from python.ebook_search.api.dependencies import (
AppConfig, # noqa: TC001 FastAPI resolves this annotated dependency at runtime
from python.ebook_search.api.dependencies import ( # noqa: TC001 FastAPI resolves these annotated dependencies at runtime
AppConfig,
AppHttpClient,
)
from python.ebook_search.bm25_corpus import bm25_index_exists, bm25_index_path, read_bm25_manifest
from python.ebook_search.llm_interface import check_chat_endpoint, check_embedding_endpoint
from python.fastapi_tools import DbSession # noqa: TC001 FastAPI resolves this annotated dependency at runtime
from python.fastapi_tools import AsyncDbSession # noqa: TC001 FastAPI resolves this annotated dependency at runtime
if TYPE_CHECKING:
from sqlalchemy.orm import Session
import httpx
from sqlalchemy.ext.asyncio import AsyncSession
from python.ebook_search.config import EbookSearchConfig
@@ -29,17 +31,17 @@ router = APIRouter()
@router.get("/health")
def health() -> dict[str, str]:
async def health() -> dict[str, str]:
"""Liveness probe that returns ok without touching dependencies."""
return {"status": "ok"}
@router.get("/ready")
def ready(config: AppConfig, session: DbSession) -> JSONResponse:
async def ready(config: AppConfig, session: AsyncDbSession, client: AppHttpClient) -> JSONResponse:
"""Readiness probe reporting database, embedding endpoint, and BM25 index status."""
database_ok = check_database(session)
embedding_ok = check_embedding_endpoint(config)
chat_status = chat_endpoint_status(config)
database_ok = await check_database(session)
embedding_ok = await check_embedding_endpoint(client, config)
chat_status = await chat_endpoint_status(client, config)
bm25_status = check_bm25_status(config)
checks = {
@@ -69,17 +71,17 @@ def ready(config: AppConfig, session: DbSession) -> JSONResponse:
return JSONResponse(content={"status": status, "checks": checks}, status_code=status_code)
def chat_endpoint_status(config: EbookSearchConfig) -> str:
async def chat_endpoint_status(client: httpx.AsyncClient, config: EbookSearchConfig) -> str:
"""Return the answering chat endpoint status, or disabled when answers are off."""
if not config.answer_enabled:
return "disabled"
return "ok" if check_chat_endpoint(config) else "fail"
return "ok" if await check_chat_endpoint(client, config) else "fail"
def check_database(session: Session) -> bool:
async def check_database(session: AsyncSession) -> bool:
"""Return whether the database answers a trivial query."""
try:
session.execute(select(literal(1)))
await session.execute(select(literal(1)))
except SQLAlchemyError as error:
logger.warning("ebook_ready_database_unavailable error=%s", error)
return False