Pyjo.JSON - Minimalistic JSON

from Pyjo.JSON import decode_json, encode_json

data = encode_json({'foo': [1, 2], 'bar': 'hello!', 'baz': True})
value = decode_json(data)

Pyjo.JSON is a minimalistic pure-Python implementation of RFC 8259.

It supports normal Python data types like dict, list, tuple, str, numbers, booleans and None, and will try to call the __json__ method on other objects, or stringify them if their class has its own __str__. Anything else becomes null. Strings and numbers are told apart by their types.

[1, -2, 3]     -> [1, -2, 3]
{"foo": "bar"} -> {'foo': 'bar'}

The keys of an object keep the order of the dict, or are sorted with sort_keys.

{'b': 1, 'a': 2} -> {"b":1,"a":2}
{'b': 1, 'a': 2} -> {"a":2,"b":1}  # sort_keys=True

Literal names will be translated to and from the native Python values, which the constants true and false hold too.

true  -> True
false -> False
null  -> None

The characters / and < will always be escaped to prevent XSS attacks, so that JSON can be put into an element script of HTML as it is.

"</script>"     -> "\u003C\/script>"
"<!--<script>"  -> "\u003C!--\u003Cscript>"

For better performance the json module of the standard library will be used, set up like the pure-Python implementation, with error messages of its own. This can also be disabled with the PYJO_NO_JSON_STDLIB environment variable.

See also

Pyjoyment, https://mojolicious.org.

Ported from Mojo::JSON of Mojolicious 9.49.

Functions

Pyjo.JSON.decode_json(data: bytes | bytearray | str) → Any[source]
value = decode_json(data)

Decode JSON to Python value and raise ValueError if decoding fails. Bytes are UTF-8, or UTF-16 or UTF-32, with a byte order mark or without one, as json.loads() detects them, where a byte order mark is not a part of the JSON.

# {'a': 1}
decode_json(b'\xff\xfe{\x00"\x00a\x00"\x00:\x001\x00}\x00')
Pyjo.JSON.default(value: object) → Any[source]
data = json.dumps(value, default=Pyjo.JSON.default)

Turn an object that is not a dict, list, tuple, str, number, boolean or None into one that is, as Pyjo.JSON does: with its __json__ method, as a string if its class has its own __str__, or as None. Bytes become a string of the characters 0-255. This is the default function for the encoder of the json module of the standard library.

# '{"url": "http://example.com"}'
json.dumps({'url': Pyjo.URL.new('http://example.com')}, default=Pyjo.JSON.default)
Pyjo.JSON.encode_json(value: object, *, sort_keys: bool = False) → bytes[source]
data = encode_json({'i': '♥ pyjoyment'})
data = encode_json({'i': '♥ pyjoyment'}, sort_keys=True)

Encode Python value to JSON, with the keys of an object in the order of the dict, or sorted with sort_keys.

Pyjo.JSON.false = False
from Pyjo.JSON import false

False value, the same as False.

Pyjo.JSON.from_json(chars: str) → Any[source]
value = from_json(chars)

Decode JSON text that is not UTF-8 encoded to Python value and raise ValueError if decoding fails.

Pyjo.JSON.j(value: list[Any] | tuple[Any, ...] | dict[Any, Any], *, sort_keys: bool = False) → bytes[source]
Pyjo.JSON.j(value: bytes | bytearray | str, *, sort_keys: bool = False) → Any
data = j([1, 2, 3])
data = j({'i': '♥ pyjoyment'})
data = j({'i': '♥ pyjoyment'}, sort_keys=True)
value = j(data)

Encode Python data structure (which may only be a list, a tuple or a dict) or decode JSON, a None return value indicates a bare null or that decoding failed. sort_keys sorts the keys of the objects of the JSON that it encodes.

Pyjo.JSON.to_json(value: object, *, sort_keys: bool = False) → str[source]
chars = to_json({'i': '♥ pyjoyment'})
chars = to_json({'i': '♥ pyjoyment'}, sort_keys=True)

Encode Python value to JSON text without UTF-8 encoding it, with the keys of an object in the order of the dict, or sorted with sort_keys.

Pyjo.JSON.true = True
from Pyjo.JSON import true

True value, the same as True.