feat: audit full alpha catalog

2026-09-07 09:37:01 +07:00
parent 600abd81f5
commit c4d1c8b87b
6 changed files with 61 additions and 6 deletions
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+from sqlalchemy import func, or_, select
+from sqlalchemy.orm import Session
+
+from .models import CatchReport, ExternalObservation, Fish, SourceType, Spot, Waterbody
+
+
+def audit_catalog(db: Session) -> dict[str, int]:
+    count = lambda model: db.scalar(select(func.count()).select_from(model)) or 0
+    failures = {
+        "invalid_weights": db.scalar(select(func.count()).select_from(CatchReport).where(or_(CatchReport.weight_g <= 0, CatchReport.weight_g > 3_000_000))) or 0,
+        "invalid_coordinates": db.scalar(select(func.count()).select_from(Spot).where(or_(Spot.x < -10_000, Spot.x > 10_000, Spot.y < -10_000, Spot.y > 10_000))) or 0,
+        "incomplete_official_records": db.scalar(select(func.count()).select_from(CatchReport).where(CatchReport.source_type == SourceType.official_record, or_(CatchReport.caught_at.is_(None), CatchReport.source_url.is_(None)))) or 0,
+        "incomplete_published_staging": db.scalar(select(func.count()).select_from(ExternalObservation).where(ExternalObservation.status == "published", or_(ExternalObservation.fish_id.is_(None), ExternalObservation.waterbody_id.is_(None), ExternalObservation.x.is_(None), ExternalObservation.y.is_(None), ExternalObservation.weight_g.is_(None), ExternalObservation.catch_report_id.is_(None)))) or 0,
+    }
+    return {"fishes": count(Fish), "waterbodies": count(Waterbody), "reports": count(CatchReport), "staging": count(ExternalObservation), **failures, "failures": sum(failures.values())}