app/document_identity.py

import hashlib
import json
from typing import Any


def content_hash(source: dict[str, Any]) -> str:
    value = {
        "article": source.get("article") or {},
        "title": source.get("title"),
        "publication_name": source.get("publication_name"),
        "publication_author": source.get("publication_author"),
        "publication_type": source.get("publication_type"),
        "source_files": source.get("source_files") or [],
    }
    return hashlib.sha256(json.dumps(value, ensure_ascii=False, sort_keys=True, separators=(",", ":")).encode()).hexdigest()