Like read_bioc_json, yielding documents one at a time.
Usage
Source
iter_bioc_json( path, *, verify="raise", offset_unit="utf8", label_key="type", concept_keys=("identifier",) )