check templates

docs/scripts/check_templates.py

@@ -17,6 +17,17 @@
 )
 CHAT_MODEL_REGEX = r".*".join(CHAT_MODEL_HEADERS)
 
+DOCUMENT_LOADER_HEADERS = (
+    "## Overview",
+    "### Integration details",
+    "### Loader features",
+    "## Setup",
+    "## Instantiation",
+    "## Load",
+    "## Lazy Load",
+    "## API reference",
+)
+DOCUMENT_LOADER_REGEX = r".*".join(DOCUMENT_LOADER_HEADERS)
 
 def check_chat_model(path: Path) -> None:
     with open(path, "r") as f:
@@ -28,13 +39,26 @@ def check_chat_model(path: Path) -> None:
             f"instructions on how to correctly format a ChatModel Integration page."
         )
 
+def check_document_loader(path: Path) -> None:
+    with open(path, "r") as f:
+        doc = f.read()
+    if not re.search(DOCUMENT_LOADER_REGEX, doc, re.DOTALL):
+        raise ValueError(
+            f"Document {path} does not match the DocumentLoader Integration page template. "
+            f"Please see https://github.com/langchain-ai/langchain/issues/22866 for "
+            f"instructions on how to correctly format a DocumentLoader Integration page."
+        )
+
 
 def main(*new_doc_paths: Union[str, Path]) -> None:
     for path in new_doc_paths:
         path = Path(path).resolve().absolute()
         if CURR_DIR.parent / "docs" / "integrations" / "chat" in path.parents:
             print(f"Checking chat model page {path}")
             check_chat_model(path)
+        elif CURR_DIR.parent / "docs" / "integrations" / "document_loaders" in path.parents:
+            print(f"Checking document loader page {path}")
+            check_document_loader(path)
         else:
             continue