add chunker.py

chunker.py

@@ -1,17 +1,19 @@
 """
-Chunking utilities for the RAG agent.
+Chunking utilities using RecursiveCharacterTextSplitter.
-
-Uses RecursiveCharacterTextSplitter from langchain.text_splitter to split documents into
-chunks of 500 characters with an overlap of 100 characters. The splitter is exposed as a
-singleton instance so that it can be reused across the project.
 """
 
 from langchain_text_splitters import RecursiveCharacterTextSplitter
 
-# Singleton splitter configuration
+# Configure splitter: chunk size 500, overlap 100
-CHUNKER = RecursiveCharacterTextSplitter(
+splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=100)
-    chunk_size=500,
-    chunk_overlap=100,
-)
 
-__all__ = ["CHUNKER"]
+def split_text(text: str) -> list[str]:
+    """Split a large text into chunks.
+
+    Args:
+        text: The raw document content.
+
+    Returns:
+        List of string chunks.
+    """
+    return splitter.split_text(text)