pyjson5.Json5IllegalCharacter in list within dictionary, within list

[lemonJumps]

Hi,

I've got an error when parsing a jsonc file, it seems to be a bug within the parser. Full error message is:

  File "src/_legacy.pyx", line 53, in pyjson5.load
  File "src/_exports.pyx", line 286, in pyjson5.decode_io
  File "src/_decoder.pyx", line 913, in pyjson5._decode_callback
  File "src/_decoder.pyx", line 815, in pyjson5._decode_all
pyjson5.Json5IllegalCharacter: ("Expected b'comma' or U+007d near 464, found U+0022", {'object': {'version': 'version', 'name': 'string', 'vertex': {'position': ['array', {'x': 'float32', 'y': 'float32', 'z': 'float32'}], 'indicies': ['array', 'optional', 'uint32'], 'normal': ['array', 'optional', {'i': 'float32', 'j': 'float32'}], 'color': ['array', 'optional', {'r': 'uint8', 'g': 'uint8', 'b': 'uint8'}]}}}, '"')

this issue happens when parsing this file:

{
    "object":
    {
        "version":"version",
        "name":"string",
        // per vertex data
        "vertex":
            {
                "position":["array", {"x":"float32", "y":"float32", "z":"float32"}],
                "indicies":["array", "optional", "uint32"],
                "normal":["array", "optional", {"i":"float32", "j":"float32"}],
                "color":["array", "optional", {"r":"uint8", "g":"uint8", "b":"uint8"}],
                "test":["array", "optional", {"test":["array", "uint8"]}]
            },
        // bundled data contains stuff links to jpegs, and other models
        // for instance having a damage model or bounding boxes
        "bundled_data":["array", "optional", "string"],
        // any would contain everything you want,
        // some custom info about the model, like wight and stuff
        "metadata":["any", "optional"]
    },
    // contains info about how the file is supposed to look like
    // type can be json, binary or text
    // json can be defined only for the entire file
    // binary and text can be selected for each parameter explicitly
    "_decorator":
    {
        // type sets, these are decoration definitions per file, and represent an arbitrary name
        // having multiple means you're expecting multiple different file types
        "json":{
            // json can't be modified as it simply dumps the file into a json
            "_type":"json"
        },        
        "ascii":{
            // data in text files will be in human readable format
            "_type":"text",
            "_structure":[
                {"name":"string", "sep":" "}, // separator is the character at the end
                {"data":"", "sep":"\n"}
            ]
        },
        "bin":{
            // data in binary will be in binary format
            "_type":"binary",
            "_hints":{
                "position":[{"position":"uint64"}], // in file position, this can only be in bytes
                "indicies":[{"position":"uint64"}],
                "normal":[{"position":"uint64"}],
                "color":[{"position":"uint64"}]
            },
            "_structure":[
                // if the size is after data the data will always be 0
                // since we need to know how big the data is before reading it
                {"name":"string[32]"}, // [xxx] < means it's size limited
                {"size":"uint64", "unit":"type"}, // units can be either type or bytes
                {"data":""} // type should be same as data that were specifien
            ]
        }
    }
}

adding new line around text fixes this:

                "test":["array", "optional", {
            "test":["array", "uint8"]
            }]

[dpranke]

Hi! Thanks for filing a bug, but I think you've got the wrong json5 library. I think you want https://github.com/Kijewski/pyjson5 instead.

[dpranke]

Closing this as it wasn't a bug in this library.