NeuroJSON Toolchain

The NeuroJSON project directly capitalizes upon the internationally standardized JSON format as the underlying data exchange format. This means all data files produced and disseminated by the project are syntactically compatible with the JSON format (or binary JSON, which can be converted to JSON). As a result, the rich set of utilities, parsers and JSON-based sub-standards, such as JSON Path, JSONLD, JSON schema etc, that are freely available in the large JSON ecosystem are readily available to processing and transforming NeuroJSON data files.

JSON-encoded data file parsers

• Python:
  • json module (built-in, no installation needed)
  • `jdata` module (https://pypi.org/project/jdata/, handles NeuroJSON data annotations)
  • orjson module (https://pypi.org/project/orjson, very fast)
  • pysimdjson module (https://pypi.org/project/pysimdjson, extremely fast)
• MATLAB:
  • jsonencode/jsondecode (built-in, MATLAB R2016 or later)
  • `jsonlab` (https://github.com/NeuroJSON/jsonlab, handles NeuroJSON data annotations)
• GNU Octave:
  • jsonencode/jsondecode (built-in, Octave 7.0 or later)
  • [[Page/jsonlab|jsonlab] (https://github.com/NeuroJSON/jsonlab, handles NeuroJSON data annotations)
• JavaScript:
  • JSON.parse/JSON.stringify (built-in)
  • jda (https://www.npmjs.com/package/jda, npm install jda, handles NeuroJSON data annotations)
• Perl:
  • use JSON::PP qw(encode_json decode_json); (built-in JSON reader/writer)
  • use JSON::XS qw(encode_json decode_json); (fast JSON reader/writer, need to install)
• C++:
  • JSON for Modern C++ (https://json.nlohmann.me/, single header, intuitive, widely used, support binary JSON)
  • simdjson (https://github.com/simdjson/simdjson, parallel JSON parser, extremely fast)
  • rapidjson (https://rapidjson.org/, fast JSON library for C++)
  • comparisons between different C/C++ JSON parsers, full list: https://github.com/miloyip/nativejson-benchmark
• C:
  • cJSON (https://github.com/DaveGamble/cJSON)
• Object Pascal/Lazarus:
  • fpjson, jsonparser (built-in)
• For other languages, see https://www.json.org/json-en.html#:~:text=%270009%27%20ws-,8th,-json

BJData/UBJSON (a quasi-human-readable binary JSON, .jdb,.bnii,.bmsh,.bnirs,.ubj,.pmat) file parsers

• Python:
  • `jdata.loadjd()` (https://pypi.org/project/jdata/, pip install jdata)
  • `iso2mesh.loadnifti()` (https://pypi.org/project/iso2mesh/, pip install iso2mesh)
• MATLAB/Octave:
  • `loadjd()/savejd()` (https://github.com/NeuroJSON/jsonlab)
  • `loadjd()/savejd()` (https://github.com/NeuroJSON/jsonlab, containing jsonlab)

NIfTI/JNIfTI/Analyze (.nii/.nii.gz/.img/.hdr) file reader/writer

• Python:
  • `jdata.loadnifti()` (https://pypi.org/project/jdata/, pip install jdata)
  • `iso2mesh.loadnifti()` (https://pypi.org/project/iso2mesh/, pip install iso2mesh)
• MATLAB/Octave:
  • `readnifti()` (https://github.com/NeuroJSON/jnifty)
  • `loadnifti()/loadjnifti()` (https://github.com/NeuroJSON/jnifty)
  • `loadnifti()/loadjnifti()` (https://github.com/NeuroJSON/jsonlab, containing jnifty)
  • `loadnifti()/loadjnifti()` (https://github.com/NeuroJSON/jsonlab, containing jsonlab)

HDF5/SNIRF (.h5/.snirf) file reader/writer

• Python:
  • `jdata.loadh5()/jdata.loadsnirf()` (https://pypi.org/project/jdata/, pip install jdata)
  • `iso2mesh.loadh5/loadsnirf/loadjsnirf` (https://pypi.org/project/iso2mesh/, pip install iso2mesh)
• MATLAB:
  • `loadh5()` (https://github.com/NeuroJSON/easyh5)
  • `loadsnirf()` (https://github.com/NeuroJSON/jsnirfy)
  • `loadsnirf()` (https://github.com/NeuroJSON/jsonlab, containing jsnirfy)
  • `loadsnirf()` (https://github.com/NeuroJSON/jsonlab, containing jsonlab)

CSV/TSV (.tsv/.csv/.tsv.gz/.csv.gz) file reader/writer

• Python:
  • `jdata.load csv tsv()/jdata.save csv tsv()` (https://pypi.org/project/jdata/, pip install jdata)
• MATLAB:
  • `loadjd()/savejd()` (https://github.com/NeuroJSON/jsonlab, handles NeuroJSON data annotations)

Binary JSON files (.msgpk/.csv/.tsv.gz/.csv.gz) file reader/writer