Ryker_Feng 063d62c3c3
feat(persistence): support custom postgres schema (#3442)
* feat(persistence): support custom postgres schema

* fix(persistence): address CI lint/test failures and review feedback

- Map missing psycopg import to actionable POSTGRES_INSTALL guidance in
  sync/async schema-creation helpers
- Accept SQLAlchemy compound DSN schemes (postgresql+asyncpg) when
  injecting search_path, normalizing to a libpq-consumable DSN
- Guard keyword-DSN tests with importorskip so they skip without psycopg
- Set database=None in sync checkpointer none-fix test to avoid MagicMock
  backend resolution
- Apply ruff import sort and format

* fix(persistence): address pg-schema review feedback

- Restrict postgres_schema regex to lowercase-only so the quoted CREATE
  SCHEMA matches the unquoted search_path (PG case-folds it), fixing the
  mixed-case bug where tables silently fell back to public.
- Replace shlex.join/split with libpq-correct backslash escaping for the
  options parameter so values containing spaces survive intact.
- Add normalize_libpq_dsn() and route the async checkpointer pool through
  dsn_with_search_path() so a +asyncpg suffix is stripped and existing DSN
  options (e.g. statement_timeout) are merged instead of overridden.
- Extract shared ensure_postgres_schema()/ensure_postgres_schema_async()
  helpers (mapping missing psycopg to the install hint) used by all four
  provider sites.
- Tests: reject mixed-case schemas, preserve space-containing libpq option,
  cover normalize_libpq_dsn, and assert pool search_path via DSN.

* fix(persistence): align pg-schema test with merged store API

The main merge moved the sync Store factory to the single-path
_resolve_store_config/_sync_store_cm design, dropping the PR's
_sync_store_from_database helper. The integration test still imported
the removed symbol, breaking test collection (backend-unit-tests).
Resolve the store config from a DatabaseConfig and drive it through
_sync_store_cm instead.

* fix(persistence): address pg-schema review feedback

- reject trailing/leading whitespace in postgres_schema via re.fullmatch
  (a $-anchored re.match let "deerflow\n" through, silently landing tables
  in public)
- re-escape all whitespace (TAB/CR/LF) when re-joining libpq options so a
  caller's pre-existing options value round-trips losslessly
- re-validate the identifier inside create_schema_sql as defense-in-depth
  at the SQL-emitting boundary
- accept the postgres:// short scheme in the alembic search_path injection
- close the sync psycopg connection explicitly (psycopg3 __exit__ does not
  close()), mirroring the async path
- drop the partial checkpointer/store reset on a database config change;
  database is restart-required and the ORM engine is not rebuilt, so a
  partial reset would half-migrate the deployment

* docs(config): complete the postgres_schema migration checklist

Address PR review (P1): the documented `public`->schema migration only
moved runs, run_events, threads_meta, feedback, and users. That strands
every other DeerFlow-owned table -- the four channel_* tables, both
scheduled_* tables, agents, and (critically) alembic_version -- in
`public`. On restart bootstrap treats the partially-populated target
schema as unversioned, re-baselines it, and replays migrations while the
real rows stay invisible in `public`.

List the full owned set explicitly, call out alembic_version as required,
and keep the "discover the rest" query for version-drift safety.

* refactor(checkpointer): drop test-only _sync_checkpointer_from_database

Address PR review: the helper was only reached by the env-gated
integration test and re-implemented the DatabaseConfig->CheckpointerConfig
backend resolution that _resolve_checkpointer_config already owns, so a
future backend added there would silently miss this path. Mirror the store
side of the same test, which reuses the production path directly:
_resolve_checkpointer_config(...) + _sync_checkpointer_cm(...).
2026-07-30 13:51:17 +08:00

226 lines
8.4 KiB
Python

"""Sync Store factory.
Provides a **sync singleton** and a **sync context manager** for CLI tools
and the embedded :class:`~deerflow.client.DeerFlowClient`.
The deprecated ``checkpointer`` section takes precedence when present;
otherwise Store follows the unified ``database`` section. Supported backends:
memory, sqlite, postgres.
Usage::
from deerflow.runtime.store.provider import get_store, store_context
# Singleton — reused across calls, closed on process exit
store = get_store()
# One-shot — fresh connection, closed on block exit
with store_context() as store:
store.put(("ns",), "key", {"value": 1})
"""
from __future__ import annotations
import contextlib
import logging
import threading
from collections.abc import Iterator
from langgraph.store.base import BaseStore
from deerflow.config.app_config import AppConfig, get_app_config
from deerflow.config.checkpointer_config import CheckpointerConfig, ensure_config_loaded, get_checkpointer_config
from deerflow.persistence.postgres_schema import dsn_with_search_path, ensure_postgres_schema
from deerflow.runtime.store._sqlite_utils import ensure_sqlite_parent_dir, resolve_sqlite_conn_str
logger = logging.getLogger(__name__)
# ---------------------------------------------------------------------------
# Error message constants
# ---------------------------------------------------------------------------
SQLITE_STORE_INSTALL = "langgraph-checkpoint-sqlite is required for the SQLite store. Install it with: uv add langgraph-checkpoint-sqlite"
POSTGRES_STORE_INSTALL = (
"langgraph-checkpoint-postgres is required for the PostgreSQL store. Install the package extra with: pip install 'deerflow-harness[postgres]' (or use: uv sync --all-packages --extra postgres when developing locally)"
)
POSTGRES_CONN_REQUIRED = "checkpointer.connection_string is required for the postgres backend"
def _ensure_postgres_schema(conn_string: str, schema: str) -> None:
"""Create the configured schema before LangGraph creates its store tables."""
ensure_postgres_schema(conn_string, schema, install_hint=POSTGRES_STORE_INSTALL)
def _resolve_store_config(app_config: AppConfig) -> CheckpointerConfig:
"""Resolve the Store backend from legacy or unified application config.
The legacy ``checkpointer`` section remains authoritative when present so
Store and Checkpointer continue to use the same backend. Otherwise the
unified ``database`` section drives the Store as documented. The unified
``postgres_schema`` is forwarded so Store tables land in the configured
schema alongside the checkpointer and app tables.
"""
if app_config.checkpointer is not None:
return app_config.checkpointer
database = app_config.database
if database is None or database.backend == "memory":
return CheckpointerConfig(type="memory")
if database.backend == "sqlite":
return CheckpointerConfig(type="sqlite", connection_string=database.checkpointer_sqlite_path)
if database.backend == "postgres":
if not database.postgres_url:
raise ValueError("database.postgres_url is required for the postgres backend")
return CheckpointerConfig(type="postgres", connection_string=database.postgres_url, postgres_schema=database.postgres_schema)
raise ValueError(f"Unknown database backend: {database.backend!r}")
def _get_store_config() -> CheckpointerConfig:
"""Load Store config without holding the provider singleton lock."""
ensure_config_loaded()
# Preserve callers that initialise the legacy config singleton directly.
legacy_config = get_checkpointer_config()
if legacy_config is not None:
return legacy_config
try:
app_config = get_app_config()
except FileNotFoundError:
return CheckpointerConfig(type="memory")
return _resolve_store_config(app_config)
# ---------------------------------------------------------------------------
# Sync factory
# ---------------------------------------------------------------------------
@contextlib.contextmanager
def _sync_store_cm(config) -> Iterator[BaseStore]:
"""Context manager that creates and tears down a sync Store.
The ``config`` argument is a
:class:`~deerflow.config.checkpointer_config.CheckpointerConfig` instance —
the same object used by the checkpointer factory.
"""
if config.type == "memory":
from langgraph.store.memory import InMemoryStore
logger.info("Store: using InMemoryStore (in-process, not persistent)")
yield InMemoryStore()
return
if config.type == "sqlite":
try:
from langgraph.store.sqlite import SqliteStore
except ImportError as exc:
raise ImportError(SQLITE_STORE_INSTALL) from exc
conn_str = resolve_sqlite_conn_str(config.connection_string or "store.db")
ensure_sqlite_parent_dir(conn_str)
with SqliteStore.from_conn_string(conn_str) as store:
store.setup()
logger.info("Store: using SqliteStore (%s)", conn_str)
yield store
return
if config.type == "postgres":
try:
from langgraph.store.postgres import PostgresStore # type: ignore[import]
except ImportError as exc:
raise ImportError(POSTGRES_STORE_INSTALL) from exc
if not config.connection_string:
raise ValueError(POSTGRES_CONN_REQUIRED)
_ensure_postgres_schema(config.connection_string, config.postgres_schema)
conn_string = dsn_with_search_path(config.connection_string, config.postgres_schema)
with PostgresStore.from_conn_string(conn_string) as store:
store.setup()
logger.info("Store: using PostgresStore")
yield store
return
raise ValueError(f"Unknown store backend type: {config.type!r}")
# ---------------------------------------------------------------------------
# Sync singleton
# ---------------------------------------------------------------------------
_store: BaseStore | None = None
_store_ctx = None # open context manager keeping the connection alive
_store_lock = threading.Lock()
def get_store() -> BaseStore:
"""Return the global sync Store singleton, creating it on first call.
The legacy ``checkpointer`` section takes precedence when configured;
otherwise the unified ``database`` section selects the backend.
Raises:
ImportError: If the required package for the configured backend is not installed.
ValueError: If the selected backend is missing its required connection value.
"""
global _store, _store_ctx
if _store is not None:
return _store
# Config loading can reset both persistence singletons. Resolve the full
# config outside this provider lock to avoid lock-order inversion.
config = _get_store_config()
with _store_lock:
if _store is not None:
return _store
store_ctx = _sync_store_cm(config)
store = store_ctx.__enter__()
_store_ctx = store_ctx
_store = store
return _store
def reset_store() -> None:
"""Reset the sync singleton, forcing recreation on the next call.
Closes any open backend connections and clears the cached instance.
Useful in tests or after a configuration change.
"""
global _store, _store_ctx
with _store_lock:
if _store_ctx is not None:
try:
_store_ctx.__exit__(None, None, None)
except Exception:
logger.warning("Error during store cleanup", exc_info=True)
_store_ctx = None
_store = None
# ---------------------------------------------------------------------------
# Sync context manager
# ---------------------------------------------------------------------------
@contextlib.contextmanager
def store_context() -> Iterator[BaseStore]:
"""Sync context manager that yields a Store and cleans up on exit.
Unlike :func:`get_store`, this does **not** cache the instance — each
``with`` block creates and destroys its own connection. Use it in CLI
scripts or tests where you want deterministic cleanup::
with store_context() as store:
store.put(("threads",), thread_id, {...})
The legacy ``checkpointer`` section takes precedence when configured;
otherwise the unified ``database`` section selects the backend.
"""
config = _resolve_store_config(get_app_config())
with _sync_store_cm(config) as store:
yield store