fix: make SEC quality gating terminal-safe

alembic/versions/028_sec_filing_retry_queue.py

@@ -4,6 +4,8 @@ Revision ID: 028
 Revises: 027
 Create Date: 2026-08-03 00:00:00.000000
 """
+from datetime import date, datetime, timezone
+import json
 from typing import Sequence, Union
 
 from alembic import op
@@ -28,11 +30,118 @@ def upgrade() -> None:
         sa.Column("coregistrant_ciks_json", sa.Text(), nullable=True),
         sa.Column("first_seen_at", sa.DateTime(timezone=True), nullable=False),
         sa.Column("last_attempted_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("escalated_at", sa.DateTime(timezone=True), nullable=True),
         sa.UniqueConstraint("accession", name="uq_sec_filing_gaps_accession"),
     )
     op.create_index("ix_sec_filing_gaps_cik", "sec_filing_gaps", ["cik"])
+    _backfill_retry_queue()
 
 
 def downgrade() -> None:
     op.drop_index("ix_sec_filing_gaps_cik", table_name="sec_filing_gaps")
     op.drop_table("sec_filing_gaps")
+
+
+def _as_date(value) -> date | None:
+    if isinstance(value, datetime):
+        return value.date()
+    if isinstance(value, date):
+        return value
+    if isinstance(value, str):
+        try:
+            return date.fromisoformat(value)
+        except ValueError:
+            return None
+    return None
+
+
+def _backfill_retry_queue() -> None:
+    """Materialize pre-queue promoted gaps once; runtime never scans history."""
+    bind = op.get_bind()
+    runs = sa.table(
+        "data_import_runs",
+        sa.column("source", sa.String()),
+        sa.column("status", sa.String()),
+        sa.column("validation_json", sa.Text()),
+        sa.column("source_max_date", sa.Date()),
+        sa.column("started_at", sa.DateTime(timezone=True)),
+    )
+    snapshots = sa.table(
+        "fundamental_snapshots",
+        sa.column("cik", sa.String()),
+        sa.column("accession", sa.String()),
+        sa.column("filed_date", sa.Date()),
+    )
+    gaps = sa.table(
+        "sec_filing_gaps",
+        sa.column("cik", sa.String()),
+        sa.column("accession", sa.String()),
+        sa.column("form", sa.String()),
+        sa.column("index_date", sa.Date()),
+        sa.column("reason", sa.String()),
+        sa.column("coregistrant_ciks_json", sa.Text()),
+        sa.column("first_seen_at", sa.DateTime(timezone=True)),
+        sa.column("last_attempted_at", sa.DateTime(timezone=True)),
+        sa.column("escalated_at", sa.DateTime(timezone=True)),
+    )
+
+    snapshot_rows = bind.execute(
+        sa.select(snapshots.c.cik, snapshots.c.accession, snapshots.c.filed_date)
+    ).all()
+    resolved_accessions = {row.accession for row in snapshot_rows}
+    latest_filed_by_cik: dict[str, date] = {}
+    for row in snapshot_rows:
+        if row.filed_date is not None:
+            current = latest_filed_by_cik.get(row.cik)
+            if current is None or row.filed_date > current:
+                latest_filed_by_cik[row.cik] = row.filed_date
+
+    audit_rows = bind.execute(
+        sa.select(
+            runs.c.validation_json,
+            runs.c.source_max_date,
+            runs.c.started_at,
+        ).where(
+            runs.c.source == "sec_facts",
+            runs.c.status == "promoted",
+            runs.c.validation_json.is_not(None),
+        )
+    ).all()
+    now = datetime.now(timezone.utc)
+    candidates: dict[str, dict] = {}
+    for audit in audit_rows:
+        try:
+            summary = json.loads(audit.validation_json)
+        except (TypeError, ValueError):
+            continue
+        if not isinstance(summary, dict):
+            continue
+        for item in summary.get("missing_xbrl") or []:
+            accession = item.get("accession")
+            raw_cik = item.get("cik")
+            if not accession or raw_cik is None or accession in resolved_accessions:
+                continue
+            cik = str(raw_cik).zfill(10)
+            index_date = _as_date(item.get("index_date")) or _as_date(
+                audit.source_max_date
+            )
+            later_filed = latest_filed_by_cik.get(cik)
+            if index_date is not None and later_filed is not None and later_filed > index_date:
+                continue
+            first_seen = audit.started_at or now
+            existing = candidates.get(accession)
+            if existing is not None and existing["first_seen_at"] <= first_seen:
+                continue
+            candidates[accession] = {
+                "cik": cik,
+                "accession": accession,
+                "form": item.get("form"),
+                "index_date": index_date,
+                "reason": item.get("reason") or "not_in_companyfacts",
+                "coregistrant_ciks_json": json.dumps(item.get("coregistrants") or []),
+                "first_seen_at": first_seen,
+                "last_attempted_at": first_seen,
+                "escalated_at": None,
+            }
+    if candidates:
+        op.bulk_insert(gaps, list(candidates.values()))