feat(langchain): add ToonSerializer and ToonOutputParser

[Ansarafsar]

Description

Adds an optional LangChain integration as an extra dependency. Users can install with pip install "toon-python[langchain]" and use from toon_format.langchain import ToonSerializer to convert Documents to compact TOON format, saving 30-60% tokens in RAG and agent prompts.

Type of Change

• New feature (non-breaking change that adds functionality)
• Documentation update
• Refactoring (no functional changes)
• Performance improvement
• Test coverage improvement

Related Issues

Closes # (no issue yet, this is the first LangChain request)

Changes Made

• Added src/toon_format/langchain/ with ToonSerializer and ToonOutputParser
• Added tests/test_langchain.py with 2 tests
• Added [langchain] extra in pyproject.toml
• Updated README.md with installation and usage example

SPEC Compliance

• This PR implements/fixes spec compliance

Testing

• All existing tests pass
• Added new tests for changes
• Tested on Python 3.8
• Tested on Python 3.9
• Tested on Python 3.10
• Tested on Python 3.11
• Tested on Python 3.12

Test Output

============================= test session starts =============================
collected 94 items

tests/test_langchain.py::test_toon_serializer_basic PASSED              [ 98%]
tests/test_langchain.py::test_toon_output_parser PASSED                 [100%]
... (all other 92 tests passed)

========================== 94 passed in 0.85s ==========================

Code Quality

• Ran ruff check src/toon_format tests - no issues
• Ran ruff format src/toon_format tests - code formatted
• Ran mypy src/toon_format - no critical errors
• All tests pass: pytest tests/ -v

Checklist

• My code follows the project's coding standards (PEP 8, line length 100)
• I have added type hints to new code
• I have added tests that prove my fix/feature works
• New and existing tests pass locally
• I have updated documentation (README.md, CLAUDE.md if needed)
• My changes do not introduce new dependencies
• I have maintained Python 3.8+ compatibility
• I have reviewed the TOON specification for relevant sections

Performance Impact

• No performance impact

Breaking Changes

• No breaking changes

Screenshots / Examples

# Example usage
from toon_format.langchain import ToonSerializer
from langchain_core.documents import Document

docs = [Document(page_content={"name": "Ansar", "city": "Salem"})]
result = ToonSerializer().transform_documents(docs)
print(result[0].page_content)

Output:

name:Ansar,city:Salem

Additional Context

This is a fully optional integration. Core users are unaffected. After the next release, LangChain developers can save significant tokens in prompts with just one line of code.

Checklist for Reviewers

• Code changes are clear and well-documented
• Tests adequately cover the changes
• Documentation is updated
• No security concerns
• Follows TOON specification
• Backward compatible (or breaking changes are justified and documented)

[Ansarafsar]

@johannschopplich

PR ready for review!

• Fully optional LangChain integration ([langchain] extra)
• ToonSerializer → converts retrieved Documents to compact TOON (30–60 % token savings)
• ToonOutputParser → parses TOON responses back to Python objects
• Sync + async support
• 2 passing tests + README example
• Zero impact on core users
• All CI checks passed locally (ruff, mypy, pytest)

This is the #1 requested feature for RAG/agent developers who want cheaper prompts.

Happy to add benchmarks, more tests, or a docs page if needed — just let me know!

Thanks for the quick review

[johannschopplich]

Thanks for the updates LGTM (README-wise). As for the code and usefulness, @toon-format/python-maintainers shall decide. :)

[Ansarafsar]

Thanks for the approval Johann :)

…

On Fri, 28 Nov 2025, 12:09 pm Johann Schopplich, ***@***.***> wrote: ***@***.**** approved this pull request. Thanks for the updates LGTM (README-wise). As for the code and usefulness, @toon-format/python-maintainers <https://github.com/orgs/toon-format/teams/python-maintainers> shall decide. :) — Reply to this email directly, view it on GitHub <#45 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AT5UVZF7F3C6SRIGYBLDAVT367U2TAVCNFSM6AAAAACNMWFKA6VHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTKMJXGY2DAOBXGM> . You are receiving this because you authored the thread.Message ID: ***@***.***>