"""Export the full-corpus reference-overlap audit, independently of map sampling."""
import hashlib
import json
from pathlib import Path

ROOT = Path(__file__).resolve().parents[1]
SOURCE = ROOT.parent / 'Ouroboros_Atlas/visualization/20261107-atlas-million-r1/coverage_audit.json'

def export_coverage():
    raw = SOURCE.read_bytes()
    audit = json.loads(raw)
    counts = audit['counts']['eligible']
    data = {'population': audit['eligible'],
            'referenceMatches': counts['swissprot2024_exact_match'],
            'beyondReference': counts['no_exact_swissprot2024_match'],
            'referenceRelease': audit['reference']['release'],
            'definition': audit['definition'], 'sequenceScope': audit['sequence_scope'],
            'source': str(SOURCE.relative_to(ROOT.parent)),
            'sourceSha256': hashlib.sha256(raw).hexdigest(),
            'interpretation': 'No exact match to this snapshot does not establish absence of other annotations. Ouroboros descriptions are model-generated functional hypotheses.'}
    assert data['referenceMatches'] + data['beyondReference'] == data['population']
    (ROOT / 'public/data/atlas-coverage.json').write_text(json.dumps(data, indent=2)+'\n')
    return data

if __name__ == '__main__':
    print(json.dumps(export_coverage(), indent=2))