backend/core/grants/versioning.py

"""Wersjonowanie zmian regulaminów i terminów (Warstwa 1)."""
from __future__ import annotations

import os
from datetime import datetime, timezone
from typing import Any, Dict, List, Optional

from sqlalchemy.orm import Session

from core.grants.models import GrantVersion, HumanVerificationItem

TRACKED_FIELDS = ("deadline", "regulation_url", "precise_regulation_url", "status", "url", "official_page_url")
VERIFICATION_FIELDS = frozenset({"deadline", "regulation_url", "precise_regulation_url", "status"})


def _norm(val: Any) -> str:
    if val is None:
        return ""
    return str(val).strip()


def detect_field_changes(
    source_id: str,
    old_item: Dict[str, Any],
    new_item: Dict[str, Any],
) -> List[Dict[str, Any]]:
    changes: List[Dict[str, Any]] = []
    for field in TRACKED_FIELDS:
        old_v = _norm(old_item.get(field))
        new_v = _norm(new_item.get(field))
        if old_v == new_v:
            continue
        if not old_v and not new_v:
            continue
        summary = _summarize_change(field, old_v, new_v)
        changes.append(
            {
                "grant_source_id": source_id,
                "field_name": field,
                "old_value": old_v,
                "new_value": new_v,
                "change_summary": summary,
                "requires_verification": field in VERIFICATION_FIELDS,
            }
        )
    return changes


def _summarize_change(field: str, old_v: str, new_v: str) -> str:
    labels = {
        "deadline": "Termin naboru",
        "regulation_url": "Regulamin",
        "precise_regulation_url": "Regulamin (precyzyjny)",
        "status": "Status",
        "url": "Strona programu",
        "official_page_url": "Oficjalna strona",
    }
    label = labels.get(field, field)
    if not old_v:
        return f"{label}: ustawiono → {new_v[:120]}"
    if not new_v:
        return f"{label}: usunięto (było: {old_v[:80]})"
    return f"{label}: {old_v[:60]} → {new_v[:60]}"


def record_changes(
    db: Session,
    changes: List[Dict[str, Any]],
    *,
    source: str = "import",
) -> Dict[str, int]:
    """Zapisuje GrantVersion + opcjonalnie HumanVerificationQueue."""
    if os.environ.get("HUMAN_VERIF_QUEUE", "true").lower() == "false":
        enqueue_hvq = False
    else:
        enqueue_hvq = True

    versions = 0
    queued = 0
    for ch in changes:
        db.add(
            GrantVersion(
                grant_source_id=ch["grant_source_id"],
                field_name=ch["field_name"],
                old_value=ch.get("old_value"),
                new_value=ch.get("new_value"),
                change_summary=ch.get("change_summary"),
                source=source,
                requires_verification=ch.get("requires_verification", False),
            )
        )
        versions += 1
        if enqueue_hvq and ch.get("requires_verification"):
            db.add(
                HumanVerificationItem(
                    grant_source_id=ch["grant_source_id"],
                    change_type="field_update",
                    field_name=ch["field_name"],
                    old_value=ch.get("old_value"),
                    new_value=ch.get("new_value"),
                    summary=ch.get("change_summary"),
                    priority=8 if ch["field_name"] == "deadline" else 6,
                )
            )
            queued += 1
    return {"versions_recorded": versions, "verification_queued": queued}


def get_recent_versions(
    db: Session,
    *,
    grant_source_id: Optional[str] = None,
    limit: int = 50,
) -> List[Dict[str, Any]]:
    q = db.query(GrantVersion).order_by(GrantVersion.detected_at.desc())
    if grant_source_id:
        q = q.filter(GrantVersion.grant_source_id == grant_source_id)
    rows = q.limit(limit).all()
    return [
        {
            "id": r.id,
            "grant_source_id": r.grant_source_id,
            "field_name": r.field_name,
            "old_value": r.old_value,
            "new_value": r.new_value,
            "change_summary": r.change_summary,
            "detected_at": r.detected_at.isoformat() if r.detected_at else None,
            "source": r.source,
            "requires_verification": r.requires_verification,
        }
        for r in rows
    ]