How to load a text file in LangChain

Plain text files are often the first source material for a retrieval prototype because they carry content without parser-specific layout concerns. In LangChain, loading a .txt file into Document objects gives splitters, retrievers, and vector stores the same input shape used by other loaders.

The Python TextLoader currently lives in the langchain_community package. It reads the file into one document by default, stores the text in page_content, and records the source path in metadata so later chunks or search results can still point back to the original file.

A minimal UTF-8 check should show the loaded document count, first content line, and source metadata. Recent langchain-community releases can print a sunset warning before the script output, so treat that warning as a package-maintenance signal rather than a failed load and pin or retest the loader before using it in long-lived ingestion jobs.

1. Open an activated Python project environment.

   Related: How to create a virtual environment for LangChain

2. Install the LangChain community loader package.

   $ python3 -m pip install --upgrade langchain-community

   langchain-community provides the TextLoader import path used by the text-file loader.
   Related: How to install LangChain with pip

3. Create a small text file to load.

   $ cat > release-notes.txt <<'TXT'
   LangChain release note
   TextLoader reads local text files into Document objects.
   Each Document keeps page_content and source metadata for later splitting or retrieval.
   TXT

4. Create the loader script.

   $ cat > load_text.py <<'PY'
   from langchain_community.document_loaders import TextLoader
   
   loader = TextLoader("release-notes.txt", encoding="utf-8")
   documents = loader.load()
   
   print("documents:", len(documents))
   print("content:", documents[0].page_content.splitlines()[0])
   print("source:", documents[0].metadata["source"])
   PY

   Set encoding when the source file is known to be UTF-8 or another specific encoding. Relying on the platform default can make the same file behave differently across systems.

5. Run the loader script.

   $ python3 load_text.py
   documents: 1
   content: LangChain release note
   source: release-notes.txt

   The document count should be 1 for this single text file, and the source metadata should match the file path passed to TextLoader.

   If Python prints a langchain-community sunset warning before the output, the import still succeeded. Recheck the upstream loader package before committing the dependency to a production ingestion pipeline.

6. Remove the temporary sample files after the load check.

   $ rm release-notes.txt load_text.py