photoai/scripts/verify-index.mjs

// Phase 0-b 색인 파이프라인 헤드리스 검증.
// sql.js 스키마 + 폴더 워크 + 샘플 해시 + upsert + count 가 실제로 동작하는지 확인.
//   node scripts/verify-index.mjs [folder]
import initSqlJs from 'sql.js'
import { readdir, stat, open, mkdtemp, writeFile, rm } from 'node:fs/promises'
import { readFileSync } from 'node:fs'
import { join, extname, dirname } from 'node:path'
import { tmpdir } from 'node:os'
import { createHash } from 'node:crypto'
import { fileURLToPath } from 'node:url'

const __dirname = dirname(fileURLToPath(import.meta.url))
const IMG = new Set(['.jpg', '.jpeg', '.png', '.webp', '.mp4', '.mov'])

async function* walk(root) {
  for (const e of await readdir(root, { withFileTypes: true })) {
    const full = join(root, e.name)
    if (e.isDirectory()) yield* walk(full)
    else if (e.isFile() && IMG.has(extname(e.name).toLowerCase())) yield full
  }
}

async function contentHash(path) {
  const s = await stat(path)
  const h = createHash('sha1')
  h.update(String(s.size))
  const len = Math.min(512 * 1024, s.size)
  if (len > 0) {
    const fh = await open(path, 'r')
    try {
      const buf = Buffer.alloc(len)
      await fh.read(buf, 0, len, 0)
      h.update(buf)
    } finally {
      await fh.close()
    }
  }
  return h.digest('hex')
}

async function main() {
  // 인자 폴더 없으면 더미 이미지 2개로 임시 폴더 생성
  let folder = process.argv[2]
  let temp = null
  if (!folder) {
    temp = await mkdtemp(join(tmpdir(), 'photoai-idx-'))
    await writeFile(join(temp, 'a.jpg'), Buffer.from('dummy-image-a'))
    await writeFile(join(temp, 'b.png'), Buffer.from('dummy-image-b-different'))
    folder = temp
    console.log('테스트 폴더 생성:', folder)
  }

  const wasm = readFileSync(join(__dirname, '..', 'node_modules', 'sql.js', 'dist', 'sql-wasm.wasm'))
  const SQL = await initSqlJs({ wasmBinary: new Uint8Array(wasm).buffer })
  const db = new SQL.Database()
  db.run(`CREATE TABLE asset (
    id INTEGER PRIMARY KEY AUTOINCREMENT, contentHash TEXT UNIQUE NOT NULL,
    path TEXT, ext TEXT, sizeBytes INTEGER, mtime INTEGER, indexedAt INTEGER);
    CREATE INDEX idx_path ON asset(path);`)

  let indexed = 0
  for await (const file of walk(folder)) {
    const s = await stat(file)
    const hash = await contentHash(file)
    db.run(
      `INSERT INTO asset (contentHash,path,ext,sizeBytes,mtime,indexedAt)
       VALUES (?,?,?,?,?,?)
       ON CONFLICT(contentHash) DO UPDATE SET path=excluded.path, mtime=excluded.mtime`,
      [hash, file, extname(file).toLowerCase(), s.size, Math.floor(s.mtimeMs), Date.now()]
    )
    indexed++
  }

  const n = db.exec('SELECT COUNT(*) FROM asset')[0].values[0][0]
  const sample = db.exec('SELECT contentHash, path FROM asset LIMIT 2')
  console.log(`색인 처리: ${indexed}건, DB asset 수: ${n}`)
  if (sample[0]) for (const row of sample[0].values) console.log('  row:', row[0].slice(0, 12), '…', row[1])

  // 재실행 시 중복 안 늘어나는지(upsert) 확인
  for await (const file of walk(folder)) {
    const s = await stat(file)
    const hash = await contentHash(file)
    db.run(
      `INSERT INTO asset (contentHash,path,ext,sizeBytes,mtime,indexedAt)
       VALUES (?,?,?,?,?,?) ON CONFLICT(contentHash) DO UPDATE SET mtime=excluded.mtime`,
      [hash, file, extname(file).toLowerCase(), s.size, Math.floor(s.mtimeMs), Date.now()]
    )
  }
  const n2 = db.exec('SELECT COUNT(*) FROM asset')[0].values[0][0]
  console.log(`재실행 후 asset 수(중복 없어야 함): ${n2}`)
  console.log(n === n2 ? 'PASS: upsert 중복 없음' : 'FAIL: 중복 발생')

  db.close()
  if (temp) await rm(temp, { recursive: true, force: true })
}

main().catch((e) => {
  console.error('오류:', e)
  process.exit(1)
})