hyperjson
hyperjson copied to clipboard
🐍 A hyper-fast Python module for reading/writing JSON data using Rust's serde-json.
We should use [smallvec](https://github.com/servo/rust-smallvec) for serializing/deserializing as it has the ability to use the stack for small vector allocations. Noticed that [orjson is doing the same](https://github.com/ijl/orjson/blob/8eea4a301d9b8d8fe726fb3bde026b30ef538257/src/decode.rs#L157-L170).
[simdjson-rs](https://github.com/simd-lite/simdjson-rs) is a Rust port of the extremely fast simdjson JSON parser with serde compatibility. We could make use of that to improve serialization/deserialization performance. This would have to be...
See #8
I've ported [CPython](https://github.com/python/cpython) 3.6.5's test cases for json module to hyperjson. I disable the `test_recursion.py` due to it crash the process. Except the `test_recursion.py`, I get `44 failed, 27 passed`...
This PR goes on top of https://github.com/mre/hyperjson/pull/43 Changes: * Initialize `entites` with default capacity > 0. This may safe some extra memory allocations. Benchmarks are again slightly contradictional. Please consider...
Bumps [simplejson](https://github.com/simplejson/simplejson) from 3.17.2 to 3.18.1. Release notes Sourced from simplejson's releases. v3.18.1 Version 3.18.1 released 2023-01-03 Remove unnecessary i variable from encoder module namespace simplejson/simplejson#303 Declare support for Python...