|
| 1 | +"""FAIR related functionalities.""" |
| 2 | + |
| 3 | +# Copyright 2025, European Union. |
| 4 | +# |
| 5 | +# Licensed under the Apache License, Version 2.0 (the "License"); |
| 6 | +# you may not use this file except in compliance with the License. |
| 7 | +# You may obtain a copy of the License at |
| 8 | +# |
| 9 | +# http://www.apache.org/licenses/LICENSE-2.0 |
| 10 | +# |
| 11 | +# Unless required by applicable law or agreed to in writing, software |
| 12 | +# distributed under the License is distributed on an "AS IS" BASIS, |
| 13 | +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 14 | +# See the License for the specific language governing permissions and |
| 15 | +# limitations under the License. |
| 16 | + |
| 17 | +import datetime |
| 18 | +import os |
| 19 | + |
| 20 | +import requests |
| 21 | +import sqlalchemy as sa |
| 22 | +import structlog |
| 23 | +from cads_common import portal |
| 24 | + |
| 25 | +from cads_catalogue import database |
| 26 | + |
| 27 | +FAIR_CHEKER_USERNAME = os.getenv("FAIR_CHECKER_USERNAME", "marvel") |
| 28 | +FAIR_CHEKER_PASSWORD = os.getenv("FAIR_CHECKER_PASSWORD", "wonderwoman") |
| 29 | + |
| 30 | +logger = structlog.get_logger(__name__) |
| 31 | + |
| 32 | + |
| 33 | +def call_fair_checker(fair_checker_host: str, site_base: str, dataset_uid: str) -> dict: |
| 34 | + """Check if the FAIR checker service is reachable. |
| 35 | +
|
| 36 | + Args: |
| 37 | + fair_checker_host: Hostname (with port, optionally) of the FAIR checker service. |
| 38 | + site_base: Base URL for DSS portal. |
| 39 | + dataset_uid: Dataset unique identifier. |
| 40 | +
|
| 41 | + Returns: JSON response from the FAIR checker service. |
| 42 | + """ |
| 43 | + payload = { |
| 44 | + "object_identifier": f"{site_base}/datasets/{dataset_uid}", |
| 45 | + "metadata_service_endpoint": "", |
| 46 | + "metadata_service_type": "oai_pmh", |
| 47 | + "use_datacite": True, |
| 48 | + "use_github": False, |
| 49 | + "metric_version": "metrics_v0.8", |
| 50 | + } |
| 51 | + |
| 52 | + logger.info("Retrieving FAIR report for resource", dataset_uid=dataset_uid) |
| 53 | + response = requests.post( |
| 54 | + f"http://{fair_checker_host}/fuji/api/v1/evaluate", |
| 55 | + headers={ |
| 56 | + "accept": "application/json", |
| 57 | + "Content-Type": "application/json", |
| 58 | + }, |
| 59 | + json=payload, |
| 60 | + auth=(FAIR_CHEKER_USERNAME, FAIR_CHEKER_PASSWORD), |
| 61 | + timeout=60, |
| 62 | + ) |
| 63 | + response.raise_for_status() |
| 64 | + return response.json() |
| 65 | + |
| 66 | + |
| 67 | +def update_fair_score(session: sa.orm.Session, fair_checker_host: str) -> None: |
| 68 | + """Update the FAIR score for all resources in the catalogue. |
| 69 | +
|
| 70 | + Args: |
| 71 | + session_obj: SQLAlchemy session object. |
| 72 | + fair_checker_service_url: URL of the FAIR checker service. |
| 73 | + """ |
| 74 | + resources = session.query(database.Resource).all() |
| 75 | + for resource in resources: |
| 76 | + site_base = portal.get_site_url(resource.portal) if resource.portal else None |
| 77 | + if not site_base: |
| 78 | + logger.warning( |
| 79 | + "Cannot determine site base URL for portal", |
| 80 | + portal=resource.portal, |
| 81 | + resource_id=resource.resource_uid, |
| 82 | + ) |
| 83 | + continue |
| 84 | + dataset_uid = resource.resource_uid |
| 85 | + assert dataset_uid is not None |
| 86 | + try: |
| 87 | + result = call_fair_checker(fair_checker_host, site_base, dataset_uid) |
| 88 | + except requests.RequestException as e: |
| 89 | + logger.error( |
| 90 | + "Error connecting to FAIR checker service", |
| 91 | + error=str(e), |
| 92 | + resource_id=resource.resource_uid, |
| 93 | + ) |
| 94 | + continue |
| 95 | + resource.fair_timestamp = datetime.datetime.utcnow() |
| 96 | + try: |
| 97 | + session.execute( |
| 98 | + sa.update(database.ResourceData) |
| 99 | + .where(database.ResourceData.resource_uid == resource.resource_uid) |
| 100 | + .values(fair_data=result), |
| 101 | + ) |
| 102 | + session.add(resource) |
| 103 | + session.commit() |
| 104 | + except sa.exc.SQLAlchemyError as e: |
| 105 | + session.rollback() |
| 106 | + logger.error( |
| 107 | + "Error updating FAIR data", |
| 108 | + error=str(e), |
| 109 | + resource_id=resource.resource_uid, |
| 110 | + ) |
| 111 | + continue |
| 112 | + logger.info("Updated FAIR data", resource_id=resource.resource_uid) |
0 commit comments