"""The shared v1 inventory contract, used at live collection and import boundaries.""" import json from functools import lru_cache from pathlib import Path from jsonschema import Draft202012Validator, FormatChecker SCHEMA_PATH = Path(__file__).resolve().parents[2] / 'schemas' / 'inventory-v1.schema.json' @lru_cache(maxsize=1) def inventory_schema(): return json.loads(SCHEMA_PATH.read_text(encoding='utf-8')) def validate_inventory(data): validator = Draft202012Validator(inventory_schema(), format_checker=FormatChecker()) if not validator.is_valid(data): raise ValueError('Inventory does not match the v1 schema; required fields, types or check statuses are invalid.') return data ALLOWED = { 'roles': {'name', 'state', 'display_name', 'feature_type'}, 'services': {'name', 'display_name', 'state', 'startup', 'account', 'pid'}, 'network': {'state', 'local_address', 'local_port', 'remote_address', 'remote_port', 'pid', 'process', 'direction'}, 'tasks': {'name', 'path', 'enabled', 'account', 'triggers', 'state', 'associated_unit', 'executables'}, 'software': {'name', 'publisher', 'version', 'installation_date', 'registry_view'}, 'storage': {'type', 'name', 'path', 'capacity_gib', 'free_gib', 'filesystem', 'special', 'observed_open_files', 'observed_client_count', 'activity_source', 'activity_observed_at'}, 'certificates': {'subject', 'issuer', 'expires', 'thumbprint', 'store'}, } SERVER_FIELDS = {'hostname', 'platform', 'operating_system', 'os_version', 'os_build', 'kernel_version', 'product_type', 'domain_joined', 'last_boot', 'ip_addresses', 'network_adapters', 'domain'} def restrict_inventory_fields(data, local_source=None): """Keep only metadata fields; never serialize raw CIM/task/service objects.""" data["server"] = {key: value for key, value in data["server"].items() if key in SERVER_FIELDS} if local_source: data["server"]["collection_source"] = local_source data["inventory"] = {section: [{key: value for key, value in row.items() if key in ALLOWED[section]} for row in rows] for section, rows in data["inventory"].items()} return data