Files
dotfiles/python/ebook_search/api/main.py
T
Richie 2706c4417d 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
2026-07-24 11:38:50 -04:00

99 lines
3.4 KiB
Python

"""FastAPI HTMX app for EPUB search."""
from __future__ import annotations
import logging
from contextlib import asynccontextmanager
from typing import TYPE_CHECKING, Annotated
import httpx
import typer
import uvicorn
from fastapi import FastAPI
from fastapi.staticfiles import StaticFiles
from sqlalchemy.ext.asyncio import AsyncSession
from python.common import configure_logger
from python.ebook_search.api.bm25_tasks import cancel_bm25_refresh
from python.ebook_search.api.routes import admin_router, health_router, page_router, search_router
from python.ebook_search.api.web import STATIC_DIR
from python.ebook_search.bm25_corpus import ensure_bm25_corpus
from python.ebook_search.config import load_config
from python.ebook_search.protected_phrases.pool import shutdown_extraction_pool
from python.fastapi_tools import ZstdMiddleware
from python.orm.common import get_async_postgres_engine
if TYPE_CHECKING:
from collections.abc import AsyncIterator
logger = logging.getLogger(__name__)
@asynccontextmanager
async def lifespan(app: FastAPI) -> AsyncIterator[None]:
"""Manage application startup and shutdown resources."""
logger.info("ebook_search_startup")
config = load_config()
app.state.config = config
logger.info(
"ebook_search_config_loaded top_k=%s embedding_model=%s embedding_base_url=%s vllm_base_url=%s "
"rerank_enabled=%s phrase_matching_enabled=%s answer_enabled=%s library_paths=%s",
config.top_k,
config.embedding_model,
config.embedding_base_url,
config.vllm_base_url,
config.rerank.enabled,
config.phrase_matching_enabled,
config.answer_enabled,
len(config.library_paths),
)
if not config.library_paths:
logger.warning("ebook_search_no_library_paths_configured")
# Concurrent phrase judging opens one session per book worker on this engine, so size the pool
# to cover those plus headroom for ordinary web requests.
app.state.engine = get_async_postgres_engine(
name="RICHIE",
vector_engine=True,
pool_size=config.phrase_judge_book_workers + 10,
)
app.state.http_client = httpx.AsyncClient()
async with AsyncSession(app.state.engine, expire_on_commit=False) as session:
await ensure_bm25_corpus(session, config)
try:
yield
finally:
logger.info("ebook_search_shutdown")
cancel_bm25_refresh(app)
shutdown_extraction_pool()
await app.state.http_client.aclose()
await app.state.engine.dispose()
def create_app() -> FastAPI:
"""Create the EPUB search web app."""
app = FastAPI(title="EPUB Search", lifespan=lifespan)
app.add_middleware(ZstdMiddleware)
app.mount("/static", StaticFiles(directory=STATIC_DIR), name="static")
app.include_router(admin_router)
app.include_router(health_router)
app.include_router(page_router)
app.include_router(search_router)
return app
def serve(
host: Annotated[str, typer.Option("--host", "-h", help="Host to bind to")] = "127.0.0.1",
port: Annotated[int, typer.Option("--port", "-p", help="Port to bind to")] = 8070,
log_level: Annotated[str, typer.Option("--log-level", "-l", help="Log level")] = "INFO",
) -> None:
"""Start the EPUB search server."""
configure_logger(log_level)
uvicorn.run(create_app(), host=host, port=port)
if __name__ == "__main__":
typer.run(serve)