lib/source_lineage.py

from __future__ import annotations

import json
from pathlib import Path
from typing import Any


def load_registry(root: str | Path) -> dict[str, Any]:
    path = Path(root) / 'lineage' / 'DONOR_REGISTRY.json'
    if not path.exists():
        return {'version': 'unavailable', 'donors': []}
    return json.loads(path.read_text(encoding='utf-8'))


def lineage_summary(root: str | Path) -> str:
    registry = load_registry(root)
    donors = registry.get('donors', [])
    embedded = [d for d in donors if d.get('embedded')]
    indexed = [d for d in donors if not d.get('embedded')]
    lines = [
        '# Source-Lineage Surface', '',
        f"Registry version: `{registry.get('version', 'unknown')}`", '',
        f"Embedded donor archives: **{len(embedded)}**", f"Hash-indexed heavy/private donors: **{len(indexed)}**", '',
        '## Governing law',
        '- Authority(State) must not exceed Support(Lineage).',
        '- Receipt presence is not equivalent to semantic validation.',
        '- External model review is assistance, not authority.',
        '- A repair compiler may prepare a descendant but may not silently replace its ancestor.', '',
        '## Embedded donors',
    ]
    for d in embedded:
        lines.append(f"- `{d['name']}` — {d.get('role','')} — `{d.get('sha256','')}`")
    lines.extend(['', '## Indexed but not embedded'])
    for d in indexed:
        lines.append(f"- `{d['name']}` — {d.get('reason_not_embedded','indexed only')} — `{d.get('sha256','')}`")
    return '\n'.join(lines)