From 14b07d06fa20354ffeed4b711dc4d84d9d968ea5 Mon Sep 17 00:00:00 2001
From: Nuno Campos <nuno@langchain.dev>
Date: Tue, 6 May 2025 11:12:16 -0700
Subject: [PATCH] Re-implement using sqlite

---
 .../langgraph/cache/file/__init__.py          | 87 ++++++++++++++-----
 1 file changed, 64 insertions(+), 23 deletions(-)

diff --git a/libs/checkpoint/langgraph/cache/file/__init__.py b/libs/checkpoint/langgraph/cache/file/__init__.py
index f37322ae2..080d24c66 100644
--- a/libs/checkpoint/langgraph/cache/file/__init__.py
+++ b/libs/checkpoint/langgraph/cache/file/__init__.py
@@ -2,17 +2,16 @@ from __future__ import annotations
 
 import asyncio
 import datetime
-import dbm
+import sqlite3
+import threading
 from collections.abc import Mapping, Sequence
 
-import ormsgpack
-
 from langgraph.cache.base import BaseCache
 from langgraph.checkpoint.serde.base import SerializerProtocol
 
 
 class FileCache(BaseCache):
-    """File-based cache using dbm."""
+    """File-based cache using SQLite."""
 
     def __init__(
         self,
@@ -22,20 +21,46 @@ class FileCache(BaseCache):
     ) -> None:
         """Initialize the cache with a file path."""
         super().__init__(serde=serde)
-        self._db = dbm.open(path, "c")
+        # SQLite backing store
+        self._conn = sqlite3.connect(
+            path,
+            check_same_thread=False,
+        )
+        # Serialize access to the shared connection across threads
+        self._lock = threading.RLock()
+        # Better concurrency & atomicity
+        self._conn.execute("PRAGMA journal_mode=WAL;")
+        # Schema: key -> (expiry, encoding, value)
+        self._conn.execute(
+            """CREATE TABLE IF NOT EXISTS cache (
+                key TEXT PRIMARY KEY,
+                expiry REAL,
+                encoding TEXT NOT NULL,
+                val BLOB NOT NULL
+            )"""
+        )
+        self._conn.commit()
 
     def get(self, keys: Sequence[str]) -> dict[str, bytes]:
         """Get the cached values for the given keys."""
-        now = datetime.datetime.now(datetime.timezone.utc).timestamp()
-        values: dict[str, bytes] = {}
-        for key in keys:
-            if val := self._db.get(key):
-                expiry, *data = ormsgpack.unpackb(val)
+        with self._lock, self._conn:
+            now = datetime.datetime.now(datetime.timezone.utc).timestamp()
+            if not keys:
+                return {}
+            placeholders = ",".join("?" for _ in keys)
+            cursor = self._conn.execute(
+                f"SELECT key, expiry, encoding, val FROM cache WHERE key IN ({placeholders})",
+                tuple(keys),
+            )
+            values: dict[str, bytes] = {}
+            rows = cursor.fetchall()
+            for key, expiry, encoding, raw in rows:
                 if expiry is not None and now > expiry:
-                    self._db.pop(key, None)
+                    # purge expired entry
+                    self._conn.execute("DELETE FROM cache WHERE key = ?", (key,))
                     continue
-                values[key] = self.serde.loads_typed(data)
-        return values
+                values[key] = self.serde.loads_typed((encoding, raw))
+            return values
 
     async def aget(self, keys: Sequence[str]) -> dict[str, bytes]:
         """Asynchronously get the cached values for the given keys."""
@@ -43,14 +68,19 @@ class FileCache(BaseCache):
 
     def set(self, mapping: Mapping[str, tuple[bytes, int | None]]) -> None:
         """Set the cached values for the given keys and TTLs."""
-        now = datetime.datetime.now(datetime.timezone.utc)
-        for key, (value, ttl) in mapping.items():
-            if ttl is not None:
-                delta = datetime.timedelta(seconds=ttl)
-                expiry: float | None = (now + delta).timestamp()
-            else:
-                expiry = None
-            self._db[key] = ormsgpack.packb((expiry, *self.serde.dumps_typed(value)))
+        with self._lock, self._conn:
+            now = datetime.datetime.now(datetime.timezone.utc)
+            for key, (value, ttl) in mapping.items():
+                if ttl is not None:
+                    delta = datetime.timedelta(seconds=ttl)
+                    expiry: float | None = (now + delta).timestamp()
+                else:
+                    expiry = None
+                encoding, raw = self.serde.dumps_typed(value)
+                self._conn.execute(
+                    "INSERT OR REPLACE INTO cache (key, expiry, encoding, val) VALUES (?, ?, ?, ?)",
+                    (key, expiry, encoding, raw),
+                )
 
     async def aset(self, mapping: Mapping[str, tuple[bytes, int | None]]) -> None:
         """Asynchronously set the cached values for the given keys and TTLs."""
@@ -58,9 +88,20 @@ class FileCache(BaseCache):
 
     def delete(self, keys: Sequence[str]) -> None:
         """Delete the cached values for the given keys."""
-        for key in keys:
-            self._db.pop(key, None)
+        if not keys:
+            return
+        with self._lock, self._conn:
+            placeholders = ",".join("?" for _ in keys)
+            self._conn.execute(
+                f"DELETE FROM cache WHERE key IN ({placeholders})", tuple(keys)
+            )
 
     async def adelete(self, keys: Sequence[str]) -> None:
         """Asynchronously delete the cached values for the given keys."""
         await asyncio.to_thread(self.delete, keys)
+
+    def __del__(self) -> None:
+        try:
+            self._conn.close()
+        except Exception:
+            pass