Centralize logic for loading from LangChainHub, add ability to pin dependencies (#805)

It's generally considered to be a good practice to pin dependencies to
prevent surprise breakages when a new version of a dependency is
released. This commit adds the ability to pin dependencies when loading
from LangChainHub.

Centralizing this logic and using urllib fixes an issue identified by
some windows users highlighted in this video -
https://youtu.be/aJ6IQUh8MLQ?t=537

Author: rowillia

docs/modules/agents/examples/load_from_hub.ipynb

@@ -62,13 +62,26 @@
     "self_ask_with_search.run(\"What is the hometown of the reigning men's U.S. Open champion?\")"
    ]
   },
+  {
+   "attachments": {},
+   "cell_type": "markdown",
+   "id": "3aede965",
+   "metadata": {},
+   "source": [
+    "# Pinning Dependencies\n",
+    "\n",
+    "Specific versions of LangChainHub agents can be pinned with the `lc@<ref>://` syntax."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
    "id": "e679f7b6",
    "metadata": {},
    "outputs": [],
-   "source": []
+   "source": [
+    "self_ask_with_search = initialize_agent(tools, llm, agent_path=\"lc@2826ef9e8acdf88465e1e5fc8a7bf59e0f9d0a85://agents/self-ask-with-search/agent.json\", verbose=True)"
+   ]
   }
  ],
  "metadata": {

langchain/agents/loading.py

@@ -1,11 +1,8 @@
 """Functionality for loading agents."""
 import json
-import os
-import tempfile
 from pathlib import Path
 from typing import Any, List, Optional, Union
 
-import requests
 import yaml
 
 from langchain.agents.agent import Agent
@@ -16,6 +13,7 @@
 from langchain.agents.tools import Tool
 from langchain.chains.loading import load_chain, load_chain_from_config
 from langchain.llms.base import BaseLLM
+from langchain.utilities.loading import try_load_from_hub
 
 AGENT_TO_CLASS = {
     "zero-shot-react-description": ZeroShotAgent,
@@ -81,29 +79,14 @@ def load_agent_from_config(
 
 def load_agent(path: Union[str, Path], **kwargs: Any) -> Agent:
     """Unified method for loading a agent from LangChainHub or local fs."""
-    if isinstance(path, str) and path.startswith("lc://agents"):
-        path = os.path.relpath(path, "lc://agents/")
-        return _load_from_hub(path, **kwargs)
+    if hub_result := try_load_from_hub(
+        path, _load_agent_from_file, "agents", {"json", "yaml"}
+    ):
+        return hub_result
     else:
         return _load_agent_from_file(path, **kwargs)
 
 
-def _load_from_hub(path: str, **kwargs: Any) -> Agent:
-    """Load agent from hub."""
-    suffix = path.split(".")[-1]
-    if suffix not in {"json", "yaml"}:
-        raise ValueError("Unsupported file type.")
-    full_url = URL_BASE + path
-    r = requests.get(full_url)
-    if r.status_code != 200:
-        raise ValueError(f"Could not find file at {full_url}")
-    with tempfile.TemporaryDirectory() as tmpdirname:
-        file = tmpdirname + "/agent." + suffix
-        with open(file, "wb") as f:
-            f.write(r.content)
-        return _load_agent_from_file(file, **kwargs)
-
-
 def _load_agent_from_file(file: Union[str, Path], **kwargs: Any) -> Agent:
     """Load agent from file."""
     # Convert file to Path object.

langchain/chains/loading.py

@@ -1,11 +1,8 @@
 """Functionality for loading chains."""
 import json
-import os
-import tempfile
 from pathlib import Path
 from typing import Any, Union
 
-import requests
 import yaml
 
 from langchain.chains.api.base import APIChain
@@ -27,6 +24,7 @@
 from langchain.chains.vector_db_qa.base import VectorDBQA
 from langchain.llms.loading import load_llm, load_llm_from_config
 from langchain.prompts.loading import load_prompt, load_prompt_from_config
+from langchain.utilities.loading import try_load_from_hub
 
 URL_BASE = "https://raw.githubusercontent.com/hwchase17/langchain-hub/master/chains/"
 
@@ -441,9 +439,10 @@ def load_chain_from_config(config: dict, **kwargs: Any) -> Chain:
 
 def load_chain(path: Union[str, Path], **kwargs: Any) -> Chain:
     """Unified method for loading a chain from LangChainHub or local fs."""
-    if isinstance(path, str) and path.startswith("lc://chains"):
-        path = os.path.relpath(path, "lc://chains/")
-        return _load_from_hub(path, **kwargs)
+    if hub_result := try_load_from_hub(
+        path, _load_chain_from_file, "chains", {"json", "yaml"}
+    ):
+        return hub_result
     else:
         return _load_chain_from_file(path, **kwargs)
 
@@ -466,19 +465,3 @@ def _load_chain_from_file(file: Union[str, Path], **kwargs: Any) -> Chain:
         raise ValueError("File type must be json or yaml")
     # Load the chain from the config now.
     return load_chain_from_config(config, **kwargs)
-
-
-def _load_from_hub(path: str, **kwargs: Any) -> Chain:
-    """Load chain from hub."""
-    suffix = path.split(".")[-1]
-    if suffix not in {"json", "yaml"}:
-        raise ValueError("Unsupported file type.")
-    full_url = URL_BASE + path
-    r = requests.get(full_url)
-    if r.status_code != 200:
-        raise ValueError(f"Could not find file at {full_url}")
-    with tempfile.TemporaryDirectory() as tmpdirname:
-        file = tmpdirname + "/chain." + suffix
-        with open(file, "wb") as f:
-            f.write(r.content)
-        return _load_chain_from_file(file, **kwargs)

langchain/prompts/loading.py

@@ -1,17 +1,15 @@
 """Load prompts from disk."""
 import importlib
 import json
-import os
-import tempfile
 from pathlib import Path
 from typing import Union
 
-import requests
 import yaml
 
 from langchain.prompts.base import BasePromptTemplate, RegexParser
 from langchain.prompts.few_shot import FewShotPromptTemplate
 from langchain.prompts.prompt import PromptTemplate
+from langchain.utilities.loading import try_load_from_hub
 
 URL_BASE = "https://raw.githubusercontent.com/hwchase17/langchain-hub/master/prompts/"
 
@@ -114,9 +112,10 @@ def _load_prompt(config: dict) -> PromptTemplate:
 
 def load_prompt(path: Union[str, Path]) -> BasePromptTemplate:
     """Unified method for loading a prompt from LangChainHub or local fs."""
-    if isinstance(path, str) and path.startswith("lc://prompts"):
-        path = os.path.relpath(path, "lc://prompts/")
-        return _load_from_hub(path)
+    if hub_result := try_load_from_hub(
+        path, _load_prompt_from_file, "prompts", {"py", "json", "yaml"}
+    ):
+        return hub_result
     else:
         return _load_prompt_from_file(path)
 
@@ -151,19 +150,3 @@ def _load_prompt_from_file(file: Union[str, Path]) -> BasePromptTemplate:
         raise ValueError(f"Got unsupported file type {file_path.suffix}")
     # Load the prompt from the config now.
     return load_prompt_from_config(config)
-
-
-def _load_from_hub(path: str) -> BasePromptTemplate:
-    """Load prompt from hub."""
-    suffix = path.split(".")[-1]
-    if suffix not in {"py", "json", "yaml"}:
-        raise ValueError("Unsupported file type.")
-    full_url = URL_BASE + path
-    r = requests.get(full_url)
-    if r.status_code != 200:
-        raise ValueError(f"Could not find file at {full_url}")
-    with tempfile.TemporaryDirectory() as tmpdirname:
-        file = tmpdirname + "/prompt." + suffix
-        with open(file, "wb") as f:
-            f.write(r.content)
-        return _load_prompt_from_file(file)

langchain/utilities/loading.py

@@ -0,0 +1,49 @@
+"""Utilities for loading configurations from langchian-hub."""
+
+import os
+import re
+import tempfile
+from pathlib import Path
+from typing import Any, Callable, Optional, Set, TypeVar, Union
+from urllib.parse import urljoin
+
+import requests
+
+DEFAULT_REF = os.environ.get("LANGCHAIN_HUB_DEFAULT_REF", "master")
+URL_BASE = os.environ.get(
+    "LANGCHAIN_HUB_URL_BASE",
+    "https://raw.githubusercontent.com/hwchase17/langchain-hub/{ref}/",
+)
+HUB_PATH_RE = re.compile(r"lc(?P<ref>@[^:]+)?://(?P<path>.*)")
+
+
+T = TypeVar("T")
+
+
+def try_load_from_hub(
+    path: Union[str, Path],
+    loader: Callable[[str], T],
+    valid_prefix: str,
+    valid_suffixes: Set[str],
+    **kwargs: Any,
+) -> Optional[T]:
+    """Load configuration from hub.  Returns None if path is not a hub path."""
+    if not isinstance(path, str) or not (match := HUB_PATH_RE.match(path)):
+        return None
+    ref, remote_path_str = match.groups()
+    ref = ref[1:] if ref else DEFAULT_REF
+    remote_path = Path(remote_path_str)
+    if remote_path.parts[0] != valid_prefix:
+        return None
+    if remote_path.suffix[1:] not in valid_suffixes:
+        raise ValueError("Unsupported file type.")
+
+    full_url = urljoin(URL_BASE.format(ref=ref), str(remote_path))
+    r = requests.get(full_url, timeout=5)
+    if r.status_code != 200:
+        raise ValueError(f"Could not find file at {full_url}")
+    with tempfile.TemporaryDirectory() as tmpdirname:
+        file = Path(tmpdirname) / remote_path.name
+        with open(file, "wb") as f:
+            f.write(r.content)
+        return loader(str(file), **kwargs)

poetry.lock

@@ -57,6 +57,7 @@ pytest-dotenv = "^0.5.2"
 duckdb-engine = "^0.6.6"
 pytest-watcher = "^0.2.6"
 freezegun = "^1.2.2"
+responses = "^0.22.0"
 
 [tool.poetry.group.lint.dependencies]
 flake8-docstrings = "^1.6.0"

pyproject.toml

@@ -0,0 +1 @@
+"""Tests utilities module."""