sori.studio/server/repositories/post-export-repository.js

import { getDefaultSiteSettings } from '../utils/site-settings'
import { getPostgresClient } from './postgres-client'

const EXPORT_STATUS = {
  QUEUED: 'queued',
  READY: 'ready'
}

const EXPORT_FILE_STATUS = {
  PENDING: 'pending'
}

const DEFAULT_CHUNK_SIZE = 100
const MAX_CHUNK_SIZE = 500
const DEFAULT_RETENTION_DAYS = 100

/**
 * 파일명에 쓸 수 없는 문자를 정리한다.
 * @param {string} value - 원본 문자열
 * @returns {string} 파일명 안전 문자열
 */
const sanitizeFilenameSegment = (value) => String(value || '')
  .trim()
  .replace(/[\\/:*?"<>|]+/g, '-')
  .replace(/\s+/g, '-')
  .replace(/-+/g, '-')
  .replace(/^-|-$/g, '')
  .slice(0, 80) || 'sori.studio'

/**
 * export 작업 행을 응답 구조로 변환한다.
 * @param {Object} row - DB 행
 * @returns {Object} export 작업
 */
const mapPostExportJobRow = (row) => ({
  id: row.id,
  requestedBy: row.requested_by || null,
  requestedEmail: row.requested_email || '',
  status: row.status,
  scope: row.scope,
  postCount: Number(row.post_count || 0),
  processedCount: Number(row.processed_count || 0),
  currentPartIndex: row.current_part_index ? Number(row.current_part_index) : null,
  chunkSize: Number(row.chunk_size || DEFAULT_CHUNK_SIZE),
  retentionDays: Number(row.retention_days || DEFAULT_RETENTION_DAYS),
  expiresAt: row.expires_at ? row.expires_at.toISOString() : null,
  message: row.message || '',
  progressMessage: row.progress_message || '',
  createdAt: row.created_at.toISOString(),
  updatedAt: row.updated_at.toISOString(),
  startedAt: row.started_at ? row.started_at.toISOString() : null,
  completedAt: row.completed_at ? row.completed_at.toISOString() : null,
  files: row.files || []
})

/**
 * export 분할 파일 행을 응답 구조로 변환한다.
 * @param {Object} row - DB 행
 * @returns {Object} export 분할 파일
 */
const mapPostExportFileRow = (row) => ({
  id: row.id,
  jobId: row.job_id,
  partIndex: Number(row.part_index || 0),
  postStart: Number(row.post_start || 0),
  postEnd: Number(row.post_end || 0),
  fileName: row.file_name,
  filePath: row.file_path || '',
  fileSizeBytes: Number(row.file_size_bytes || 0),
  status: row.status,
  createdAt: row.created_at.toISOString(),
  updatedAt: row.updated_at.toISOString(),
  completedAt: row.completed_at ? row.completed_at.toISOString() : null
})

/**
 * 요청된 분할 크기를 안전 범위로 보정한다.
 * @param {unknown} value - 입력 분할 크기
 * @returns {number} 보정된 분할 크기
 */
const normalizeChunkSize = (value) => {
  const parsed = Number(value)

  if (!Number.isFinite(parsed)) {
    return DEFAULT_CHUNK_SIZE
  }

  return Math.min(Math.max(Math.trunc(parsed), 1), MAX_CHUNK_SIZE)
}

/**
 * export 산출물 보존 일수를 안전 범위로 보정한다.
 * @param {unknown} value - 입력 보존 일수
 * @returns {number} 보정된 보존 일수
 */
const normalizeRetentionDays = (value) => {
  const parsed = Number(value)

  if (!Number.isFinite(parsed)) {
    return DEFAULT_RETENTION_DAYS
  }

  return Math.min(Math.max(Math.trunc(parsed), 1), DEFAULT_RETENTION_DAYS)
}

/**
 * 사이트 이름을 조회한다.
 * @param {ReturnType<import('postgres')>} sql - PostgreSQL 클라이언트
 * @returns {Promise<string>} 사이트 이름
 */
const getExportSiteName = async (sql) => {
  const rows = await sql`
    SELECT title
    FROM site_settings
    ORDER BY updated_at DESC
    LIMIT 1
  `

  return rows[0]?.title || getDefaultSiteSettings().title
}

/**
 * export 작업 목록 조회
 * @returns {Promise<Array>} export 작업 목록
 */
export const listPostExportJobs = async () => {
  const sql = getPostgresClient()

  if (!sql) {
    return []
  }

  const jobRows = await sql`
    SELECT *
    FROM post_export_jobs
    ORDER BY created_at DESC
    LIMIT 20
  `

  if (jobRows.length === 0) {
    return []
  }

  const jobIds = jobRows.map((row) => row.id)
  const fileRows = await sql`
    SELECT *
    FROM post_export_files
    WHERE job_id = ANY(${jobIds})
    ORDER BY job_id, part_index ASC
  `
  const filesByJobId = fileRows.reduce((groups, row) => {
    const key = row.job_id
    groups[key] = groups[key] || []
    groups[key].push(mapPostExportFileRow(row))
    return groups
  }, {})

  return jobRows.map((row) => mapPostExportJobRow({
    ...row,
    files: filesByJobId[row.id] || []
  }))
}

/**
 * 게시물 export 작업 요청 생성
 * @param {Object} input - 작업 요청 입력
 * @param {string} input.requestedBy - 요청 관리자 회원 ID
 * @param {string} input.requestedEmail - 요청 관리자 이메일
 * @param {'all'|'author'} [input.scope] - export 범위
 * @param {number} [input.chunkSize] - 분할당 게시물 수
 * @param {number} [input.retentionDays] - 산출물 보존 일수
 * @returns {Promise<Object>} 생성된 export 작업
 */
export const createPostExportJob = async (input) => {
  const sql = getPostgresClient()

  if (!sql) {
    throw new Error('DATABASE_REQUIRED')
  }

  const scope = input.scope === 'author' ? 'author' : 'all'
  const chunkSize = normalizeChunkSize(input.chunkSize)
  const retentionDays = normalizeRetentionDays(input.retentionDays)
  const siteName = sanitizeFilenameSegment(await getExportSiteName(sql))
  const expiresAt = new Date(Date.now() + retentionDays * 24 * 60 * 60 * 1000)

  const [createdJob] = await sql.begin(async (transaction) => {
    const [{ count }] = scope === 'author'
      ? await transaction`
        SELECT COUNT(*)::int AS count
        FROM posts
        WHERE author_id = ${input.requestedBy}
      `
      : await transaction`
        SELECT COUNT(*)::int AS count
        FROM posts
      `
    const postCount = Number(count || 0)
    const status = postCount > 0 ? EXPORT_STATUS.QUEUED : EXPORT_STATUS.READY
    const message = postCount > 0
      ? 'Export 작업이 대기열에 등록되었습니다.'
      : '내보낼 게시물이 없습니다.'
    const jobRows = await transaction`
      INSERT INTO post_export_jobs (
        requested_by,
        requested_email,
        status,
        scope,
        post_count,
        chunk_size,
        retention_days,
        expires_at,
        message,
        completed_at
      )
      VALUES (
        ${input.requestedBy},
        ${input.requestedEmail || ''},
        ${status},
        ${scope},
        ${postCount},
        ${chunkSize},
        ${retentionDays},
        ${expiresAt},
        ${message},
        ${postCount > 0 ? null : new Date()}
      )
      RETURNING *
    `
    const job = jobRows[0]

    if (postCount > 0) {
      const partCount = Math.ceil(postCount / chunkSize)
      const fileInputs = Array.from({ length: partCount }, (_, index) => {
        const postStart = index * chunkSize + 1
        const postEnd = Math.min((index + 1) * chunkSize, postCount)

        return {
          jobId: job.id,
          partIndex: index + 1,
          postStart,
          postEnd,
          fileName: `${siteName}_${postStart}-${postEnd}.zip`,
          status: EXPORT_FILE_STATUS.PENDING
        }
      })

      for (const fileInput of fileInputs) {
        await transaction`
          INSERT INTO post_export_files (
            job_id,
            part_index,
            post_start,
            post_end,
            file_name,
            status
          )
          VALUES (
            ${fileInput.jobId},
            ${fileInput.partIndex},
            ${fileInput.postStart},
            ${fileInput.postEnd},
            ${fileInput.fileName},
            ${fileInput.status}
          )
        `
      }
    }

    return [job]
  })

  return (await listPostExportJobs()).find((job) => job.id === createdJob.id) || mapPostExportJobRow({
    ...createdJob,
    files: []
  })
}