Instructions¶

These steps install the library, generate Mojo from a JSON Schema, and run the tests. The runtime needs Mojo 1.1.0. Code generation is a Mojo program in this repository. It does not shell out to another BSON library.

Install¶

Install pixi.
Clone the repository and install the Mojo 1.1.0 pin.

git clone https://github.com/leo-gan/gld-bson.git
cd gld-bson
pixi install

If pixi install returns 401 or 403 from conda.modular.com, put PREFIX_API_KEY in a file named .env in this directory and run scripts/ci-setup.sh. That file is listed in .gitignore. Do not commit it.

The published linux-64 package is on prefix.dev:

pixi add --channel https://prefix.dev/leo-gan/leo-gan mojo-bson

The package installs bson.mojoc, wire.mojoc, runtime.mojoc, schema.mojoc, and the gld-bsongen-mojo command.

Generate¶

A schema file is JSON Schema plus MongoDB bsonType names (int, long, double) and $ref entries under $defs. scripts/generate.sh reads testdata/schema/*.json and writes tests/generated/.

pixi run generate
pixi run mojo run -I src src/codegen/cli.mojo -- --help

Test¶

pixi run test
pixi run check-generated
python3 tests_interop/oracle.py

pixi run test runs every tests/test_*.mojo file. The oracle encodes documents with pymongo and checks that this library decodes and re-encodes the same bytes. pymongo is not linked into the Mojo binary.

This site¶

pip install -r requirements-docs.txt
mkdocs build --strict
mkdocs serve

The published copy is leo-gan.github.io/gld-bson.