Session 2.5: Data Management Plan & deposit

# Import os to read env vars
import os
# Import bob and helper functions
from bia_bob import bob, available_models, fix, doc

# Initialize assistant, reads API key from env var automatically
bob.initialize(
    # read endpoint URL from env var
    endpoint=os.getenv('ENDPOINT_URL'),
    # select an available model for your purpose
    model="mini",
    # read the Data Steward system prompt from env var
    system_prompt=os.getenv('SYSTEM_PROMPT_DATA_STEWARD'),
)

%%bob
Draft a Data Management Plan section for a small research dataset, following the
Horizon Europe DMP structure (1. Data summary; 2. FAIR data with subsections on findable,
accessible, interoperable, reusable; plus short notes on storage/backup and licensing).
Write it in Markdown and store it to "dmp_draft.md"

Use ONLY these facts and do not invent details:
- Dataset: biological field samples of penguins (species, island, collection date, body mass,
  flipper length, sex), ~180 records, CSV format.
- Provenance: raw CSV handed over from a research project; cleaned and validated by the data steward.
- Validation: a pandera schema enforces types, controlled vocabularies (species, island, sex),
  value ranges, and a unique sample_id. Stored as data_schema.yml.
- Quality: a profiling report was produced. Known remaining issues are documented and flagged
  back to the data provider: some missing required values and a few impossible/out-of-range
  measurements (NOT silently changed).
- Packaging: a Frictionless Data Package (datapackage.json) carries a data dictionary and metadata.
- Findability/PID: the package will be deposited to a repository (Zenodo) which mints a DOI.
- Interoperability: open CSV format, Frictionless Table Schema, controlled vocabularies.
- Licensing: CC-BY-4.0.
- Storage: kept with version control; repository handles long-term preservation.

import os
import requests

ZENODO_TOKEN = os.getenv("ZENODO_SANDBOX_TOKEN")
# SANDBOX only — not the real Zenodo
BASE = "https://sandbox.zenodo.org/api"

HAVE_TOKEN = bool(ZENODO_TOKEN)
print("Token found - deposit cells will run." if HAVE_TOKEN
      else "No ZENODO_SANDBOX_TOKEN set - deposit cells will skip. Follow the demo instead.")

if HAVE_TOKEN:
    headers = {"Authorization": f"Bearer {ZENODO_TOKEN}"}
    r = requests.post(f"{BASE}/deposit/depositions", json={}, headers=headers)
    r.raise_for_status()
    deposition = r.json()
    deposition_id = deposition["id"]
    bucket_url = deposition["links"]["bucket"]
    draft_doi = deposition["metadata"]["prereserve_doi"]["doi"]
    print("Created deposition:", deposition_id)
    print("Reserved test DOI:", draft_doi)
    print("View draft:", deposition["links"]["html"])

if HAVE_TOKEN:
    files_to_upload = [
        "../outputs/field_samples_clean.csv",
        "../outputs/datapackage.json",
    ]
    for path in files_to_upload:
        filename = os.path.basename(path)
        with open(path, "rb") as fp:
            up = requests.put(f"{bucket_url}/{filename}", data=fp, headers=headers)
        up.raise_for_status()
        print("Uploaded:", filename)

import json

if HAVE_TOKEN:
    metadata = {
        "metadata": {
            "title": "Penguin Field Samples 2026 (training deposit)",
            "upload_type": "dataset",
            "description": (
                "Cleaned and validated biological field-sample measurements, packaged as a "
                "Frictionless Data Package. Created during the EOSC SSDS 2026 training. "
                "Test deposit on the Zenodo Sandbox."
            ),
            "creators": [{"name": "Data Steward", "affiliation": "EOSC SSDS 2026"}],
            "access_right": "open",
            "license": "cc-by-4.0",
            "keywords": ["penguins", "biology", "field samples", "FAIR", "training"],
        }
    }
    meta_headers = {"Content-Type": "application/json",
                    "Authorization": f"Bearer {ZENODO_TOKEN}"}
    r = requests.put(f"{BASE}/deposit/depositions/{deposition_id}",
                     data=json.dumps(metadata), headers=meta_headers)
    r.raise_for_status()
    print("Metadata added. Review the draft in your browser before publishing:")
    print(r.json()["links"]["html"])

if HAVE_TOKEN:
    # set confirm to True only when you have reviewed the draft and want to publish
    confirm = True
    if confirm:
        r = requests.post(f"{BASE}/deposit/depositions/{deposition_id}/actions/publish",
                          headers=headers)
        r.raise_for_status()
        published = r.json()
        print("Published! Your test DOI:", published["doi"])
        print("Record:", published["links"]["record_html"])
    else:
        print("Not published - set confirm = True to publish once you have reviewed the draft.")