fix(IT): drop 88 placeholder Provincia rows (#1349 follow-up) (#1482)

* feat(postcodes/KH): bulk-import 1,640 Cambodia postcodes via Cambodia Post + fix KH regex (#1039)

Source: Cambodia Post 2017-reform 6-digit catalogue redistributed via
the seanghay/cambodia-postal-codes JSON. All 25 provinces resolve at
100% via direct numeric-iso2 lookup — the source's "id" field (1-25)
is identical to CSC's state.iso2 for Cambodia provinces. Records
dedupe at (postcode, sangkat + district) granularity.

Also fixes the Cambodia postal_code_regex/format in countries.json:
the previous "#####" / "^(\\d{5})$" never matched Cambodia Post's
post-2017 6-digit codes (e.g. 120101 for Phnom Penh / Khan Chamkar
Mon / Tonle Basak) and would have rejected every legitimate row.
Updated to "######" / "^(\\d{6})$".

Refs #1039.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* fix(IT): drop 87 placeholder Provincia rows (#1349 follow-up)

Removes 87 placeholder "Provincia ..." records (ids 59104-59190)
from contributions/cities/IT.json. These were leftover province-level
pseudo-cities from the pre-#1395 schema; after the city→province
remap, every real comune resolves directly to its province via
state_id, so the placeholders are duplicate concepts.

contributions/cities/IT.json: 9,947 → 9,860.

Refs #1349.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: dr5hn

.github/fixes-docs/FIX_1349_SUMMARY.md

@@ -148,3 +148,46 @@ The script is idempotent: a second run on already-remapped data produces 0 chang
 - Wikipedia — Provinces of Italy: https://en.wikipedia.org/wiki/Provinces_of_Italy
 - Wikipedia — Metropolitan cities of Italy: https://en.wikipedia.org/wiki/Metropolitan_cities_of_Italy
 - Predecessor fix: `.github/fixes-docs/ITALY_MISSING_METROPOLITAN_CITIES_AND_AUTONOMOUS_PROVINCES.md`
+
+---
+
+## Follow-up: drop placeholder "Provincia ..." rows (2026-04-27)
+
+The reporter on #1349 specifically called out "Provincia di Lucca don't have sense to exist". Those rows were not real comuni — they were **province-level placeholder cities** carried over from the pre-#1395 schema, when cities pointed at regions and a separate "Provincia di X" pseudo-city stood in for the province below the region. After the #1395 remap, every real comune resolves directly to its province via `state_id` / `state_code`, so the placeholders are a duplicate concept and have been removed.
+
+### Selection
+
+- **Range:** ids 59104–59190 inclusive (contiguous, no gaps).
+- **Count:** 87 rows.
+- **Names:** all start with `Provincia ` — covers the standard `Provincia di X` form plus a handful of variants (`Provincia autonoma di Trento`, `Provincia dell' Aquila`, `Provincia Verbano-Cusio-Ossola`).
+
+### Counts
+
+| | Before | After |
+|---|--:|--:|
+| `IT.json` city records | 9,947 | 9,860 |
+| Rows starting with `Provincia ` | 87 | 0 |
+
+(Note: the prompt's expected post-#1479 baseline of 9,941 was off by 6; the correct baseline at the time of this PR was 9,947, confirmed by `git log` and `jq '. \| length'`. Range arithmetic 59190 − 59104 + 1 = 87, not 88.)
+
+### Implementation
+
+`bin/scripts/fixes/italy_drop_provincia_placeholders.py` — defensive double-predicate filter (id range AND name prefix). Refuses to touch rows in the id range that don't match the name prefix (exit 2). Idempotent — second run is a no-op.
+
+### Validation
+
+- ✅ `jq '. | length'` → 9,860.
+- ✅ `jq '[.[] | select(.name | startswith("Provincia "))] | length'` → 0.
+- ✅ No `parent_id` references to the dropped id range.
+- ✅ Neighbour ids 59103 and 59191 preserved.
+- ✅ JSON parses cleanly via `python3 -m json.tool`.
+- ✅ `normalize_json.py` reports "All records already have IDs and timestamps" — no other fields touched.
+
+### Files
+
+- Modified: `contributions/cities/IT.json` (87 rows removed).
+- Added: `bin/scripts/fixes/italy_drop_provincia_placeholders.py`.
+
+### Still open on #1349
+
+Two dedup pairs from the eight flagged in the original remap remain to be merged. They are tracked separately and not closed by this PR.

bin/scripts/fixes/italy_drop_provincia_placeholders.py

@@ -0,0 +1,136 @@
+#!/usr/bin/env python3
+"""Drop placeholder "Provincia ..." rows from contributions/cities/IT.json.
+
+Issue #1349 follow-up. The reporter flagged entries like "Provincia di Lucca"
+that "don't have sense to exist" — they are not comuni, they are leftover
+province-level placeholder records from the pre-#1395 schema (cities used
+to point at regions, so a separate "Provincia di X" pseudo-city stood in
+for the province). Since #1395 (`italy_remap_cities.py`) re-parented every
+real comune onto its province via `state_id` / `state_code`, these 87
+placeholder rows are now duplicate concepts and should be removed.
+
+Selection (defensive — both predicates must hold):
+  - id in the contiguous range [59104, 59190] (87 records, no gaps).
+  - name starts with "Provincia " (covers "Provincia di X",
+    "Provincia autonoma di Trento", "Provincia dell' Aquila",
+    "Provincia Verbano-Cusio-Ossola").
+
+The script preserves every other record byte-for-byte; canonical
+re-formatting is delegated to `bin/scripts/sync/normalize_json.py`.
+
+Idempotent: a second run on already-cleaned data writes nothing and exits 0.
+
+Usage:
+    python3 bin/scripts/fixes/italy_drop_provincia_placeholders.py [--dry-run]
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import sys
+from pathlib import Path
+from typing import List
+
+REPO_ROOT = Path(__file__).resolve().parents[3]
+CITIES_JSON = REPO_ROOT / "contributions/cities/IT.json"
+
+ID_MIN = 59104
+ID_MAX = 59190
+NAME_PREFIX = "Provincia "
+EXPECTED_DROP_COUNT = 87
+EXPECTED_PRE_COUNT = 9947
+EXPECTED_POST_COUNT = EXPECTED_PRE_COUNT - EXPECTED_DROP_COUNT  # 9860
+
+
+def load_cities(path: Path) -> List[dict]:
+    """Read the cities JSON file and verify the top-level shape."""
+    with path.open("r", encoding="utf-8") as fh:
+        data = json.load(fh)
+    if not isinstance(data, list):
+        raise SystemExit(f"Expected JSON array at {path}, got {type(data).__name__}")
+    return data
+
+
+def is_placeholder(row: dict) -> bool:
+    """Return True if the row is a legacy 'Provincia ...' placeholder."""
+    rid = row.get("id")
+    name = row.get("name", "")
+    return (
+        isinstance(rid, int)
+        and ID_MIN <= rid <= ID_MAX
+        and isinstance(name, str)
+        and name.startswith(NAME_PREFIX)
+    )
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    parser.add_argument(
+        "--dry-run",
+        action="store_true",
+        help="Report what would be dropped without rewriting the file.",
+    )
+    args = parser.parse_args()
+
+    if not CITIES_JSON.exists():
+        raise SystemExit(f"Cities file not found: {CITIES_JSON}")
+
+    cities = load_cities(CITIES_JSON)
+    pre_count = len(cities)
+
+    to_drop = [row for row in cities if is_placeholder(row)]
+    kept = [row for row in cities if not is_placeholder(row)]
+
+    drop_count = len(to_drop)
+    post_count = len(kept)
+
+    # Cross-check that nothing else in the id range slipped in.
+    in_range = [row for row in cities if isinstance(row.get("id"), int)
+                and ID_MIN <= row["id"] <= ID_MAX]
+    non_placeholder_in_range = [row for row in in_range if not is_placeholder(row)]
+
+    print(f"Input file:          {CITIES_JSON.relative_to(REPO_ROOT)}")
+    print(f"Pre-clean count:     {pre_count}")
+    print(f"Rows in id range:    {len(in_range)} (ids {ID_MIN}-{ID_MAX})")
+    print(f"Placeholder matches: {drop_count}")
+    print(f"Post-clean count:    {post_count}")
+
+    if non_placeholder_in_range:
+        print("\nERROR: rows in the placeholder id range that don't match "
+              "the name prefix — refusing to touch them:", file=sys.stderr)
+        for row in non_placeholder_in_range:
+            print(f"  id={row.get('id')} name={row.get('name')!r}", file=sys.stderr)
+        return 2
+
+    if drop_count == 0:
+        print("\nNo placeholder rows found — nothing to do (idempotent).")
+        return 0
+
+    if drop_count != EXPECTED_DROP_COUNT or pre_count != EXPECTED_PRE_COUNT:
+        print(
+            f"\nWARNING: counts diverge from the expected baseline "
+            f"({EXPECTED_PRE_COUNT} → drop {EXPECTED_DROP_COUNT} → "
+            f"{EXPECTED_POST_COUNT}). Got {pre_count} → drop {drop_count} "
+            f"→ {post_count}. Review the diff before merging.",
+            file=sys.stderr,
+        )
+
+    if args.dry_run:
+        print("\n--dry-run: not writing.")
+        sample = ", ".join(repr(r["name"]) for r in to_drop[:3])
+        print(f"Sample drops: {sample}, ...")
+        return 0
+
+    with CITIES_JSON.open("w", encoding="utf-8") as fh:
+        json.dump(kept, fh, ensure_ascii=False, indent=2)
+        fh.write("\n")
+
+    print(f"\nWrote {post_count} records to {CITIES_JSON.relative_to(REPO_ROOT)}.")
+    print("Run `python3 bin/scripts/sync/normalize_json.py "
+          "contributions/cities/IT.json` next to canonicalize formatting.")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())