tree-sitter-language-pack

What it is and what it does

tree-sitter-language-pack wraps the tree-sitter parsing library and bundles pre-compiled grammars for 371 languages behind a single Python API. Instead of managing individual language grammars and compilation, you install this package and call get_language() to receive a ready-to-use parser for any supported language. Parsers are downloaded on first use and cached locally, keeping the initial install small while supporting selective language installation.

The package is designed for code analysis, inspection, and extraction tasks: pulling out functions, classes, imports, and symbols from source; feeding code into RAG pipelines; or giving language models structural understanding of a codebase. It bundles standard tree-sitter query types (highlights, injections, locals, tags, indents, folds) per language and maintains a compiled-query cache. The single runtime dependency is tree-sitter itself, and it supports Python 3.10 through 3.14.

Use it for:

• Extract functions, classes, and imports from source files for code indexing or documentation generation
• Prepare code chunks for retrieval-augmented generation (RAG) systems feeding language models
• Build developer tooling that needs to understand and navigate syntax trees across multiple languages
• Analyze and split large codebases into LLM-digestible pieces for code review or refactoring agents
• Implement syntax-aware code transformations or linting rules without writing language-specific parsers

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides pre-compiled tree-sitter parsers for 371 programming languages, enabling fast syntax tree parsing and code analysis without requiring grammar compilation.

Yes. The package is actively maintained (released 9 days ago), has no known vulnerabilities, uses a permissive MIT license, and solves a real problem—parsing 371 languages without compilation overhead. Medium install friction is acceptable given the breadth of language support and the lean dependency graph. Reach for it whenever you need to process or analyze code structurally.

Install

tree-sitter-language-pack on PyPI

pip install tree-sitter-language-pack

uv add tree-sitter-language-pack

poetry add tree-sitter-language-pack

Installing tree-sitter-language-pack

Before you install

Medium install friction due to platform-specific wheels (macOS, Linux, Windows, ARM variants), but wheels are pre-built and the package has active maintenance with a recent release (9 days old). Single runtime dependency on tree-sitter keeps the dependency graph lean.

License in practice

MIT license is permissive; you can use this package freely in commercial and private projects without restriction or obligation to share modifications.

Quickstart

pip install tree-sitter-language-pack

from tree_sitter_language_pack import get_language

lang = get_language('python')
parser = lang.parser()
tree = parser.parse(b'def hello(): pass')

Requires Python 3.10 or later; parsers are downloaded on first use and cached locally.

Verify before relying

• Whether on-demand parser downloads work offline after initial cache population
• Performance characteristics (parse speed, memory overhead) compared to direct tree-sitter usage
• Whether all 371 languages have equivalent feature parity or if some have limited query support

Package facts

Evidence: tree_sitter_language_pack-1.14.3-cp310-abi3-macosx_10_12_x86_64.whl; tree_sitter_language_pack-1.14.3-cp310-abi3-macosx_11_0_arm64.whl; tree_sitter_language_pack-1.14.3-cp310-abi3-manylinux_2_34_aarch64.whl; tree_sitter_language_pack-1.14.3-cp310-abi3-manylinux_2_34_x86_64.whl; tree_sitter_language_pack-1.14.3-cp310-abi3-win_amd64.whl; tree_sitter_language_pack-1.14.3-cp310-abi3-win_arm64.whl

Keywords: language-pack, parser, syntax, tree-sitter

Tags

More Linguistic packages