datalad_next.itertools.load_json

datalad_next.itertools.load_json(iterable: Iterable[bytes | str]) → Generator[Any, None, None]

Convert items yielded by iterable into JSON objects and yield them

This function fetches items from the underlying iterable. The items are expected to be bytes, str, or bytearry, and contain one JSON-encoded object. Items are converted into a JSON-object, by feeding them into json.loads.

On successful conversion to a JSON-object, load_json will yield the resulting JSON-object. If the conversion to a JSON-object fails, load_json will raise a json.decoder.JSONDecodeError:

>>> from datalad_next.itertools import load_json, load_json_with_flag
>>> tuple(load_json(['{"a": 1}']))
({'a': 1},)
>>> tuple(load_json(['{"c": 3']))   # Faulty JSON-encoding, doctest: +SKIP
Traceback (most recent call last):
    ...
json.decoder.JSONDecodeError: Expecting ',' delimiter: line 1 column 8 (char 7)

Using load_json together with itemize allows the processing of JSON-lines data. itemize will yield a single item for each line and load_json will convert it into a JSON-object.

Note: JSON-decoding is slightly faster if the items of type str. Items of type bytes or bytearray will work as well, but processing might be slower.

Parameters:

iterable (Iterable[bytes | str]) -- The iterable that yields the JSON-strings or -bytestrings that should be parsed and converted into JSON-objects

Yields:

Any -- The JSON-object that are generated from the data yielded by iterable

Raises:

json.decoder.JSONDecodeError -- If the data yielded by iterable is not a valid JSON-string