pyjson5

A Python implementation of the JSON5 data format.

JSON5 extends the JSON data interchange format to make it slightly more usable as a configuration language:

JavaScript-style comments (both single and multi-line) are legal.
Object keys may be unquoted if they are legal ECMAScript identifiers
Objects and arrays may end with trailing commas.
Strings can be single-quoted, and multi-line string literals are allowed.

There are a few other more minor extensions to JSON; see the above page for the full details.

This project implements a reader and writer implementation for Python; where possible, it mirrors the standard Python JSON API package for ease of use.

There is one notable difference from the JSON api: the load() and loads() methods support optionally checking for (and rejecting) duplicate object keys; pass allow_duplicate_keys=False to do so (duplicates are allowed by default).

This is an early release. It has been reasonably well-tested, but it is SLOW. It can be 1000-6000x slower than the C-optimized JSON module, and is 200x slower (or more) than the pure Python JSON module.

Please Note: This library only handles JSON5 documents, it does not allow you to read arbitrary JavaScript. For example, bare integers can be legal object keys in JavaScript, but they aren't in JSON5.

Known issues

Did I mention that it is SLOW?
The implementation follows Python3's json implementation where possible. This means that the encoding method to dump() is ignored, and unicode strings are always returned.
The cls keyword argument that json.load()/json.loads() accepts to specify a custom subclass of JSONDecoder is not and will not be supported, because this implementation uses a completely different approach to parsing strings and doesn't have anything like the JSONDecoder class.
The cls keyword argument that json.dump()/json.dumps() accepts is also not supported, for consistency with json5.load(). The default keyword is supported, though, and might be able to serve as a workaround.

Contributing

json5 have no runtime dependencies and it is supported on Python version 3.8 or later. However, in order to develop and build the package you need a bunch of extra tools and the latest versions of those tools may require 3.9 or later. You can install the extra environment on 3.8 (and get older versions of the tools), but they may not run completely cleanly.

On Mac

The easiest thing to do is to install uv and use uv and the //run script to develop things. See ./run --help for the various commands that are supported. glop is the parser generator tool used to generate a parser from the grammar in json5/json5.g.

$ brew install uv
$ git clone https://github.com/dpranke/pyjson5
$ git clone https://github.com/dpranke/glop
$ cd pyjson5
$ source $(./run devenv)  # To activate a venv w/ all the needed dev tools.

On other platforms

Install uv via whatever mechanism is appropriate.

Create the venv

$ ./run devenv

(This calls uv sync --extra dev.)

Running the tests

$ ./run tests

Updating the packages

# Update the version in json5/version.py to $VERSION, which should be of
# the form X.Y.Z where X, Y, and Z are numbers.
$ ./run regen
$ ./run presubmit
$ git commit -a -m "Bump the version to $VERSION"
$ git tag "v$VERSION"
$ ./run build
$ ./run publish --prod
$ git push origin
$ git push --tags origin

(Assuming you have upload privileges to PyPI and the GitHub repo, of course.)

Version History / Release Notes

v0.13.0 (2026-01-01)
- No code changes.
- Add Python 3.14 to supported version, project config, dependencies
- Update dependencies to latest stuff < 2025-12-01
  - This relaxes the versions specified in pyproject.toml to just use 'newer than' rather than exact matches.
  - Sets 'uv.tool.exclude-newer' in pyproject.toml to tell uv not to look at packages published within the past 30 days; this will hopefully help prevent dependencies on compromised projects.
- Switch to using dependency-groups for 'dev' group.
v0.12.1 (2025-08-12)
- Fix #94, where objects returned from a custom encoder were not being indented properly.
v0.12.0 (2025-04-03)
- Roll back pyproject.toml change for licensing so that we can still build the package on 3.8.
- Upgrade devenv package dependencies to latest versions; they now need Python 3.9 or newer, though json5 itself still supports 3.8.
v0.11.0 (2025-04-01)
- Add a couple examples to the documentation and run doctest over them.
- Fix a typing issue in dump and dumps with the cls argument; turns out mypy was right and I was wrong and I didn't realize it :).
- Introduce a new parse method that can be used to iterate through a string, extracting multiple values.
- Add a new consume_trailing parameter to load/loads/parse that specifies whether to keep parsing after a valid object is reached. By default, this is True and the string must only contain trailing whitespace. If set to False, parsing will stop when a valid object is reached.
- Add a new start parameter to load/loads/parse to specify the zero-based offset to start parsing the string or file from.
- GitHub issue #60. Fix a bug where we were attempting to allow '--4' as a valid number.
v0.10.0 (2024-11-25)
- GitHub issue #57. Added a JSON5Encoder class that can be overridden to do custom encoding of values. This class is vaguely similar to the JSONEncoder class in the standard json library, except that it has an encode() method that can be overridden to customize any value, not just ones the standard encoder doesn't know how to handle. It does also support a default() method that can be used to encode things not normally encodable, like the JSONEncoder class. It does not support an iterencode method. One could probably be added in the future, although exactly how that would work and interact with encode is a little unclear.
- Restructured the code to use the new encoder class; doing so actually allowed me to delete a bunch of tediously duplicative code.
- Added a new quote_style argument to dump()/dumps() to control how strings are encoded by default. For compatibility with older versions of the json5 library and the standard json library, it uses QuoteStyle.ALWAYS_DOUBLE which encodes all strings with double quotes all the time. You can also configure it to use single quotes all the time (ALWAYS_SINGLE), and to switch between single and double when doing so eliminates a need to escape quotes (PREFER_SINGLE and PREFER_DOUBLE). This also adds a --quote-style argument to python -m json5.
- This release has a fair number of changes, but is intended to be completely backwards-compatible. Code without changes should run exactly as it did before.
v0.9.28 (2024-11-11)
- Fix GitHub CI to install uv so ./run tests works properly.
- Mark Python3.13 as supported in package metadata.
- Update dev package dependencies (note that the latest versions of coverage and pylint no longer work w/ Python3.8)
v0.9.27 (2024-11-10)
- Fix typo in //README.md
v0.9.26 (2024-11-10)
- GitHub issue #82 Add support for the strict parameter to load()/loads().
- Significantly rework the infra and the run script to be contemporary.
v0.9.25 (2024-04-12)
- GitHub issue #81 Explicitly specify the directory to use for the package in pyproject.toml.
v0.9.24 (2024-03-16)
- Update GitHub workflow config to remove unnecessary steps and run on pull requests as well as commits.
- Added note about removing hypothesize in v0.9.23.
- No code changes.
v0.9.23 (2024-03-16)
- Lots of cleanup:
  - Removed old code needed for Python2 compatibility.
  - Removed tests using hypothesize. This ran model-based checks and didn't really add anything useful in terms of coverage to the test suite, and it introduced dependencies and slowed down the tests significantly. It was a good experiment but I think we're better off without it.
  - Got everything linting cleanly with pylint 3.1 and ruff check using ruff 0.3.3 (Note that commit message in 00d73a3 says pylint 3.11, which is a typo).
  - Code reformatted with ruff format
  - Added missing tests to bring coverage up to 100%.
  - Lots of minor code changes as the result of linting and coverage testing, but no intentional functional differences.
v0.9.22 (2024-03-06)
- Attempt to fix the GitHub CI configuration now that setup.py is gone. Also, test on 3.12 instead of 3.11.
- No code changes.
v0.9.21 (2024-03-06)
- Moved the benchmarks/*.json data files' license information to //LICENSE to (hopefully) make the Google linter happy.
v0.9.20 (2024-03-03)
- Added json5.__version__ in addition to json5.VERSION.
- More packaging modernization (no more setup.{cfg,py} files).
- Mark Python3.12 as supported in project.classifiers.
- Updated the //run script to use python3.
v0.9.19 (2024-03-03)
- Replaced the benchmarking data files that came from chromium.org with three files obtained from other datasets on GitHub. Since this re

json5

Description