Add checkup review flags, duplicate detection, and corrected painting images.

Introduce checkup_checked/checkup_fixed on paintings with Checkup page filters and API. Fixed paintings auto-mark as reviewed. Add find-duplicates tooling and document debug/checkup workflow. Include Botticelli and Michelangelo image fixes from checkup. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-06-20 20:45:55 +03:00
parent 4683089495
commit 4c6acd5a3a
55 changed files with 750 additions and 39 deletions
@@ -0,0 +1,124 @@
+require('dotenv').config();
+const pool = require('../server/db');
+
+async function main() {
+  const exact = await pool.query(`
+    SELECT a.name AS artist, p.title, p.year, COUNT(*)::int AS cnt,
+           array_agg(p.id ORDER BY p.id) AS ids,
+           array_agg(COALESCE(p.image_path, '') ORDER BY p.id) AS image_paths,
+           array_agg(COALESCE(p.wikipedia_title, '') ORDER BY p.id) AS wiki_titles
+    FROM paintings p
+    JOIN artists a ON a.id = p.artist_id
+    GROUP BY a.id, a.name, p.title, p.year
+    HAVING COUNT(*) > 1
+    ORDER BY cnt DESC, a.name, p.title
+  `);
+
+  const normalized = await pool.query(`
+    SELECT a.name AS artist,
+           lower(regexp_replace(regexp_replace(p.title, '[^a-zA-Z0-9 ]', '', 'g'), '\\s+', ' ', 'g')) AS norm_title,
+           COUNT(*)::int AS cnt,
+           array_agg(p.id ORDER BY p.id) AS ids,
+           array_agg(p.title ORDER BY p.id) AS titles,
+           array_agg(COALESCE(p.year::text, '?') ORDER BY p.id) AS years
+    FROM paintings p
+    JOIN artists a ON a.id = p.artist_id
+    GROUP BY a.id, a.name, norm_title
+    HAVING COUNT(*) > 1
+    ORDER BY cnt DESC, a.name, norm_title
+  `);
+
+  const rublev = await pool.query(`
+    SELECT p.id, p.title, p.year, p.image_path, p.wikipedia_title,
+           p.checkup_checked, p.checkup_fixed
+    FROM paintings p
+    JOIN artists a ON a.id = p.artist_id
+    WHERE a.name ILIKE '%rublev%'
+       OR p.title ILIKE '%trinity%'
+    ORDER BY a.name, p.title, p.id
+  `);
+
+  const stats = await pool.query(`
+    SELECT COUNT(*)::int AS dup_groups,
+           COALESCE(SUM(cnt - 1), 0)::int AS extra_rows
+    FROM (
+      SELECT COUNT(*)::int AS cnt
+      FROM paintings
+      GROUP BY artist_id, title, year
+      HAVING COUNT(*) > 1
+    ) d
+  `);
+
+  const byArtist = await pool.query(`
+    SELECT a.name, COUNT(*)::int AS dup_groups, SUM(cnt - 1)::int AS extra_rows
+    FROM (
+      SELECT artist_id, title, year, COUNT(*)::int AS cnt
+      FROM paintings GROUP BY artist_id, title, year HAVING COUNT(*) > 1
+    ) d
+    JOIN artists a ON a.id = d.artist_id
+    GROUP BY a.id, a.name ORDER BY extra_rows DESC
+  `);
+
+  const rublevCluster = await pool.query(`
+    SELECT p.id, p.title, p.year, p.wikipedia_title, p.image_path,
+           (SELECT COUNT(*) FROM painting_influences pi
+            WHERE pi.painting_id = p.id OR pi.influenced_by_painting_id = p.id) AS influence_edges,
+           (SELECT COUNT(*) FROM painting_influence_sources pis
+            WHERE pis.painting_id = p.id OR pis.source_painting_id = p.id) AS source_edges
+    FROM paintings p
+    JOIN artists a ON a.id = p.artist_id
+    WHERE a.name ILIKE '%rublev%'
+      AND (p.title ILIKE '%trinity%' OR p.title ILIKE '%hospitality%')
+    ORDER BY p.id
+  `);
+
+  const total = await pool.query('SELECT COUNT(*)::int AS n FROM paintings');
+
+  console.log('=== Summary ===');
+  console.log(`Total paintings: ${total.rows[0].n}`);
+  console.log(`Exact duplicate groups: ${stats.rows[0].dup_groups} (${stats.rows[0].extra_rows} rows could be removed)`);
+  console.log('\nArtists with exact duplicates:');
+  for (const row of byArtist.rows) {
+    console.log(`  ${row.extra_rows} extra — ${row.name} (${row.dup_groups} duplicated works)`);
+  }
+
+  console.log('\n=== Exact duplicates (same artist + title + year) ===');
+  console.log(`Groups: ${exact.rows.length}`);
+  for (const row of exact.rows) {
+    console.log(`\n[${row.cnt}x] ${row.artist} — "${row.title}" (${row.year ?? '?'})`);
+    console.log(`  ids: ${row.ids.join(', ')}`);
+    console.log(`  images: ${row.image_paths.join(' | ')}`);
+    if (row.wiki_titles.some(Boolean)) console.log(`  wiki: ${row.wiki_titles.join(' | ')}`);
+  }
+
+  console.log('\n=== Normalized title duplicates (same artist, similar title) ===');
+  console.log(`Groups: ${normalized.rows.length}`);
+  for (const row of normalized.rows) {
+    const alreadyExact = exact.rows.some(
+      (e) => e.artist === row.artist && e.cnt === row.cnt && e.ids.join() === row.ids.join()
+    );
+    if (alreadyExact) continue;
+    console.log(`\n[${row.cnt}x] ${row.artist}`);
+    row.titles.forEach((t, i) => console.log(`  id ${row.ids[i]}: "${t}" (${row.years[i]})`));
+  }
+
+  console.log('\n=== Rublev Trinity / Hospitality cluster (likely same icon) ===');
+  for (const row of rublevCluster.rows) {
+    console.log(`  id ${row.id}: "${row.title}" (${row.year ?? '?'})`);
+    console.log(`    wiki: ${row.wikipedia_title || '—'}`);
+    console.log(`    image: ${row.image_path || '—'}`);
+    console.log(`    influence links: ${row.influence_edges} legacy, ${row.source_edges} sources`);
+  }
+
+  console.log('\n=== All Rublev / Trinity title matches ===');
+  for (const row of rublev.rows) {
+    console.log(`  id ${row.id}: "${row.title}" (${row.year ?? '?'}) — ${row.image_path || 'no image'}`);
+  }
+
+  await pool.end();
+}
+
+main().catch((err) => {
+  console.error(err);
+  process.exit(1);
+});