Upgrade tenacity to respect min time for exponential backoff

Fix for issue is in tenacity 9.0.0. But older langchain required
tenacity <0.9.0.

Explicitly pin version of langchain sub packages to avoid indexing
and doc parsing breakage.

pyproject.toml

@@ -44,7 +44,7 @@ dependencies = [
     "jinja2 == 3.1.6",
     "openai >= 1.0.0",
     "tiktoken >= 0.3.2",
-    "tenacity >= 8.2.2",
+    "tenacity >= 9.0.0",
     "magika ~= 0.5.1",
     "pillow ~= 10.0.0",
     "pydantic[email] >= 2.0.0",
@@ -57,10 +57,9 @@ dependencies = [
     "torch == 2.6.0",
     "uvicorn == 0.30.6",
     "aiohttp ~= 3.9.0",
-    "langchain == 0.2.5",
-    "langchain-community == 0.2.5",
+    "langchain-text-splitters == 0.3.1",
+    "langchain-community == 0.3.3",
     "requests >= 2.26.0",
-    "tenacity == 8.3.0",
     "anyio ~= 4.8.0",
     "pymupdf == 1.24.11",
     "django == 5.1.8",

src/khoj/processor/content/text_to_entries.py

@@ -6,7 +6,7 @@ from abc import ABC, abstractmethod
 from itertools import repeat
 from typing import Any, Callable, List, Set, Tuple
 
-from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_text_splitters import RecursiveCharacterTextSplitter
 from tqdm import tqdm
 
 from khoj.database.adapters import (

src/khoj/processor/conversation/anthropic/anthropic_chat.py

@@ -4,7 +4,7 @@ from datetime import datetime, timedelta
 from typing import AsyncGenerator, Dict, List, Optional
 
 import pyjson5
-from langchain.schema import ChatMessage
+from langchain_core.messages.chat import ChatMessage
 
 from khoj.database.models import Agent, ChatModel, KhojUser
 from khoj.processor.conversation import prompts

src/khoj/processor/conversation/anthropic/utils.py

@@ -3,7 +3,7 @@ from time import perf_counter
 from typing import Dict, List
 
 import anthropic
-from langchain.schema import ChatMessage
+from langchain_core.messages.chat import ChatMessage
 from tenacity import (
     before_sleep_log,
     retry,

src/khoj/processor/conversation/google/gemini_chat.py

@@ -4,7 +4,7 @@ from datetime import datetime, timedelta
 from typing import AsyncGenerator, Dict, List, Optional
 
 import pyjson5
-from langchain.schema import ChatMessage
+from langchain_core.messages.chat import ChatMessage
 from pydantic import BaseModel, Field
 
 from khoj.database.models import Agent, ChatModel, KhojUser

src/khoj/processor/conversation/google/utils.py

@@ -9,7 +9,7 @@ import httpx
 from google import genai
 from google.genai import errors as gerrors
 from google.genai import types as gtypes
-from langchain.schema import ChatMessage
+from langchain_core.messages.chat import ChatMessage
 from pydantic import BaseModel
 from tenacity import (
     before_sleep_log,

src/khoj/processor/conversation/offline/chat_model.py

@@ -7,7 +7,7 @@ from time import perf_counter
 from typing import Any, AsyncGenerator, Dict, List, Optional, Union
 
 import pyjson5
-from langchain.schema import ChatMessage
+from langchain_core.messages.chat import ChatMessage
 from llama_cpp import Llama
 
 from khoj.database.models import Agent, ChatModel, KhojUser

src/khoj/processor/conversation/openai/gpt.py

@@ -4,7 +4,7 @@ from datetime import datetime, timedelta
 from typing import AsyncGenerator, Dict, List, Optional
 
 import pyjson5
-from langchain.schema import ChatMessage
+from langchain_core.messages.chat import ChatMessage
 from openai.lib._pydantic import _ensure_strict_json_schema
 from pydantic import BaseModel
 

src/khoj/processor/conversation/prompts.py

@@ -1,4 +1,4 @@
-from langchain.prompts import PromptTemplate
+from langchain_core.prompts import PromptTemplate
 
 ## Personality
 ## --

src/khoj/processor/conversation/utils.py

@@ -17,7 +17,7 @@ import pyjson5
 import requests
 import tiktoken
 import yaml
-from langchain.schema import ChatMessage
+from langchain_core.messages.chat import ChatMessage
 from llama_cpp import LlamaTokenizer
 from llama_cpp.llama import Llama
 from transformers import AutoTokenizer, PreTrainedTokenizer, PreTrainedTokenizerFast

tests/test_conversation_utils.py

@@ -1,7 +1,7 @@
 from copy import deepcopy
 
 import tiktoken
-from langchain.schema import ChatMessage
+from langchain_core.messages.chat import ChatMessage
 
 from khoj.processor.conversation import utils