Changelog

This changelog is generated automatically from GitHub Releases.

v0.5.7

2026-06-17 · GitHub

[0.5.7] - 2026-06-17

Added

  • guess_extension(content_type, with_dot=True) — reverse lookup from a MIME type to its canonical file extension (e.g. application/pdf → .pdf), the inverse of get_content_type(). Case-insensitive, and MIME parameters on a Content-Type header are ignored (text/html; charset=utf-8 → .html). Common non-canonical / legacy spellings also resolve (e.g. text/json → .json, image/jpg → .jpg, application/x-zip-compressed → .zip). Unknown types return None. (#9)
  • guess_all_extensions(content_type, with_dot=True) — returns every known extension for a MIME type, canonical first (e.g. image/jpeg → ['.jpg', '.jpeg', '.jpe']); unknown types return []. Same case-insensitivity, parameter-stripping, and alias handling as guess_extension(). (#9)
  • Files: content_types/__init__.py, tests/test_content_types.py, README.md, great-docs.yml, docs/

Changed

  • get_content_type() now distinguishes an omitted fallback from an explicit fallback=None. Passing fallback=None returns None for unknown extensions (handy when you want to branch on a miss rather than receive a placeholder type), while omitting fallback is unchanged and still returns application/octet-stream (or text/plain with treat_as_binary=False) — so existing callers are unaffected. The default keyword switched from None to a private sentinel to tell the two cases apart, and the return type widened to Optional[str]. (#8)
  • Files: content_types/__init__.py, tests/test_content_types.py, README.md

v0.4.0

2026-06-11 · GitHub

[0.4.0] - 2026-06-10

Added

  • fallback parameter on get_content_type() to override the default MIME type returned for unknown extensions — pass any string and it will be returned in place of application/octet-stream / text/plain (#8)
  • URL support: get_content_type() now handles query strings and URL fragments
  • MIME types for .sas (text/x-sas) and .sql (application/sql) extensions (#5)
  • Comprehensive test suite (35 tests) covering integration, spot tests, and negative conditions
  • Documentation site built with Great Docs and published at https://mkennedy.codes/docs/content-types/ — auto-generated API reference plus a README-driven landing page. Built locally with scripts/build_docs.py and served from the committed docs/ folder (scripts/serve_docs.py previews it under the production subpath).
  • Files: content_types/__init__.py, tests/test_content_types.py, tests/__init__.py, pytest.ini, great-docs.yml, scripts/build_docs.py, scripts/serve_docs.py, pyproject.toml, docs/

Changed

  • get_content_type() now raises TypeError (instead of a bare Exception) when called with None. The message is unchanged, so existing except Exception handlers keep working.
  • Modernized type hints (dict[str, str], Optional[str]) and rewrote the get_content_type() / cli() docstrings in Google style (Args/Returns/Raises) for a cleaner generated API reference.
  • .yaml, .yml, and .kubeconfig now map to application/yaml (RFC 9512, registered 2024) instead of the legacy text/yaml. The content_types.yaml shortcut constant updates to match.
  • Added the Typing :: Typed trove classifier to advertise the existing PEP 561 py.typed marker on PyPI; added a test asserting the marker ships with the package.
  • Dynamic version metadata: package now pulls __version__ from distribution metadata.
  • The sdist now contains only the package, change log, readme, and license — previously it bundled every git-tracked file (the built docs/ site, tests/, scripts/, editor config, etc.). Wheels were already minimal. ([tool.hatch.build.targets.sdist] in pyproject.toml)
  • Files: content_types/__init__.py, pyproject.toml, README.md, tests/test_content_types.py

Fixed

  • .ass (Advanced SubStation Alpha subtitles) now maps to text/x-ssa — matching its sibling .ssa — instead of the incorrect audio/aac. .rst (reStructuredText, value unchanged at text/x-rst) moved out of the audio block into the documentation block.
  • Docstring example for get_content_type("script.js") now shows 'text/javascript' (the actual return value) instead of the outdated 'application/javascript'
  • Files: content_types/__init__.py

v0.3.0

2025-10-01 · GitHub

Added

  • 137 new file extensions across 17 categories, expanding format recognition capabilities
  • Comprehensive support for data-science MIME types (e.g., application/vnd.pandas, application/x-ipynb+json)
  • Project now supports 360+ file formats
  • Warp Project Summary / Index document for contributors and users
  • Files: content_types/__init__.py, WARP.md

Changed

  • Implemented alphabetical sorting of output listings for better navigation
  • Enhanced README to highlight 360+ supported file formats
  • Improved docstrings throughout codebase for clarity
  • Files: content_types/__init__.py, README.md

Fixed

  • Corrected CLI help text instructions for clearer usage guidance
  • Adjusted code indentation for better visual consistency
  • Files: content_types/__init__.py

v0.2.3

2025-02-01 · GitHub

  • Changes .js back to text/javascript
  • Adds a few new types
  • Adds comparison to builtin mimetypes

v0.2.2

2025-01-31 · GitHub

Added py.typed file to suppress mypy typing warnings. Thanks @sanders41 Available on PyPI now.

v0.2.1

2025-01-31 · GitHub

Adds many more file extensions as known types.

v0.2.0

2025-01-31 · GitHub

Initial public release.