schema_pattern_linting.py

import click
import pprint

from linkml_runtime import SchemaView
from linkml_runtime.dumpers import yaml_dumper
from linkml_runtime.utils.schemaview import OrderedBy
from linkml_runtime.linkml_model.meta import EnumDefinition


@click.command()
@click.option('--schema-file', default='src/schema/nmdc.yaml', help='Path to schema file.')
def main(schema_file):
    mixs_source_path = "https://raw.githubusercontent.com/microbiomedata/nmdc-schema/main/src/schema/mixs.yaml"
    schema_view = SchemaView(schema_file)

    schema_slots = schema_view.all_slots(ordered_by=OrderedBy.LEXICAL)

    print("Report of slots that aren't associated with any classes:\n")
    for sk, sv in schema_slots.items():
        classes_for_slot = schema_view.get_classes_by_slot(sv)
        if len(classes_for_slot) == 0:
            slot_uri_suffix = f", {sv.slot_uri}" if sv.slot_uri else ""
            slot_children = schema_view.slot_children(sk)
            slot_children_prefix = f"parent " if len(slot_children) > 0 else ""
            print(f"No classes for {slot_children_prefix}slot {sk}{slot_uri_suffix}")
    print("\n")

    print("Report of enums that aren't associated with any slots:\n")
    schema_enums = schema_view.all_enums()
    for ek, ev in schema_enums.items():
        slots_for_enum = schema_view.get_slots_by_enum(ek)
        if len(slots_for_enum) == 0:
            print(f"No slots for {ek}")
    print("\n")

    print("Report of all types. Manual review recommended:\n")
    schema_types = schema_view.all_types()
    pprint.pprint(list(schema_types.keys()))
    print("\n")

    schema_elements = schema_view.all_elements()

    whitespace_keys = [key for key in schema_elements.keys() if any(char.isspace() for char in key)]

    print("Report of all elements whose names contain whitespace:\n")
    for wsk in whitespace_keys:
        wse = schema_view.get_element(wsk)

        if type(wse) is EnumDefinition:
            wse.permissible_values = None

        if wse.from_schema != mixs_source_path:
            print(yaml_dumper.dumps(wse))
    print("\n")

    print("Report of subsets usage:\n")
    declared_subsets = []
    subsets = schema_view.all_subsets()
    for sbk, sbv in subsets.items():
        if sbv.from_schema != mixs_source_path:
            declared_subsets.append(sbk)

    subset_usage = {}
    for ds in declared_subsets:
        subset_usage[ds] = []

    for ek, ev in schema_elements.items():
        if ev.in_subset:
            for subset in ev.in_subset:
                if subset in declared_subsets:
                    subset_usage[subset].append(ev.name)

    pprint.pprint(subset_usage)


if __name__ == '__main__':
    main()