HIPCTF2/backend/src/modules/admin/system/backup.service.ts

import { Injectable, Logger } from '@nestjs/common';
import { ConfigService } from '@nestjs/config';
import { DataSource } from 'typeorm';
import * as fs from 'fs';
import * as path from 'path';
import * as crypto from 'crypto';
import {
  ADMIN_SYSTEM_BACKUP_FORMAT,
  ADMIN_SYSTEM_BACKUP_VERSION,
} from './dto/admin-system.dto';
import { ApiError } from '../../../common/errors/api-error';
import { ERROR_CODES } from '../../../common/errors/error-codes';

const INTERNAL_TABLES = new Set<string>([
  'admin_operation_token',
  'migrations',
  'sqlite_sequence',
  'sqlite_master',
  'sqlite_temp_master',
  'sqlite_temp_schema',
]);

export interface BackupFileEntry {
  path: string;
  originalFilename?: string;
  mimeType?: string;
  sizeBytes: number;
  sha256: string;
  dataBase64: string;
}

export interface BackupObject {
  format: typeof ADMIN_SYSTEM_BACKUP_FORMAT;
  version: number;
  exportedAt: string;
  tables: Record<string, Array<Record<string, unknown>>>;
  tableCounts: Record<string, number>;
  uploads: BackupFileEntry[];
}

/**
 * Generates a complete database + uploads backup as a single in-memory
 * JSON object. The resulting payload is intentionally unversioned at the
 * application level beyond the project's backup format identifier — new
 * tables are captured automatically because they are discovered from
 * `sqlite_master` at generation time rather than being hard-coded.
 */
@Injectable()
export class BackupService {
  private readonly logger = new Logger(BackupService.name);
  private readonly uploadDir: string;

  constructor(config: ConfigService, private readonly dataSource: DataSource) {
    this.uploadDir = path.resolve(config.get<string>('UPLOAD_DIR', './data/uploads'));
  }

  /**
   * Build the complete backup. Any failure during table snapshot or file
   * reading throws without partial results — the caller is expected to
   * translate this into an `SYSTEM_BACKUP_FAILED` envelope.
   */
  async build(): Promise<BackupObject> {
    const exportedAt = new Date().toISOString();
    const tables: Record<string, Array<Record<string, unknown>>> = {};
    const tableCounts: Record<string, number> = {};
    const tablesList = await this.discoverTables();
    for (const tableName of tablesList) {
      const rows = await this.safeRead(tableName);
      tables[tableName] = rows.map((row) => ({ ...row }));
      tableCounts[tableName] = rows.length;
    }
    const uploads = await this.collectUploads();
    return {
      format: ADMIN_SYSTEM_BACKUP_FORMAT,
      version: ADMIN_SYSTEM_BACKUP_VERSION,
      exportedAt,
      tables,
      tableCounts,
      uploads,
    };
  }

  /**
   * Returns every application table — dynamic discovery from the live
   * SQLite schema. Internal metadata tables and operational tables that
   * must not be restored are excluded.
   */
  async discoverTables(): Promise<string[]> {
    const rows = await this.safeQuery<{ name: string }>(
      `SELECT name FROM sqlite_master WHERE type='table' ORDER BY name ASC`,
    );
    return rows.map((r) => r.name).filter((n) => !INTERNAL_TABLES.has(n));
  }

  /**
   * Convenience for tests: report the upload root resolved at construction time.
   */
  getUploadDir(): string {
    return this.uploadDir;
  }

  private async safeRead(table: string): Promise<Record<string, unknown>[]> {
    if (!/^[a-zA-Z_][a-zA-Z0-9_]*$/.test(table)) {
      throw new ApiError(ERROR_CODES.SYSTEM_BACKUP_FAILED, `Invalid table name: ${table}`, 500);
    }
    const rows = await this.safeQuery<Record<string, unknown>>(`SELECT * FROM "${table}"`);
    return rows.map((row) => {
      const out: Record<string, unknown> = {};
      for (const [k, v] of Object.entries(row)) {
        if (v instanceof Buffer) {
          // better-sqlite3 returns BLOB columns as Buffers; encode so JSON
          // round-trips without losing data.
          out[k] = { __blob_b64: v.toString('base64') };
        } else {
          out[k] = v;
        }
      }
      return out;
    });
  }

  private async safeQuery<T>(sql: string): Promise<T[]> {
    try {
      return (await this.dataSource.query(sql)) as T[];
    } catch (err) {
      throw new ApiError(
        ERROR_CODES.SYSTEM_BACKUP_FAILED,
        `Backup query failed: ${(err as Error)?.message ?? String(err)}`,
        500,
        [{ path: 'database', message: sql }],
      );
    }
  }

  private async collectUploads(): Promise<BackupFileEntry[]> {
    const out: BackupFileEntry[] = [];
    if (!fs.existsSync(this.uploadDir)) return out;
    const visit = (dir: string) => {
      const entries = fs.readdirSync(dir, { withFileTypes: true });
      for (const entry of entries) {
        if (entry.name === '.staging') continue;
        const target = path.join(dir, entry.name);
        if (entry.isDirectory()) {
          visit(target);
        } else if (entry.isFile()) {
          const stat = fs.statSync(target);
          const buf = fs.readFileSync(target);
          out.push({
            path: path.relative(this.uploadDir, target).split(path.sep).join('/'),
            originalFilename: entry.name,
            sizeBytes: stat.size,
            sha256: crypto.createHash('sha256').update(buf).digest('hex'),
            dataBase64: buf.toString('base64'),
          });
        }
      }
    };
    try {
      visit(this.uploadDir);
    } catch (err) {
      throw new ApiError(
        ERROR_CODES.SYSTEM_FILE_READ_FAILED,
        `Failed to read uploads: ${(err as Error)?.message ?? String(err)}`,
        500,
      );
    }
    return out;
  }

  /**
   * Best-effort JSON stringifier used by the controller when streaming the
   * download. Centralized so tests can build deterministic snapshots.
   */
  static stringify(obj: BackupObject): string {
    return JSON.stringify(obj);
  }
}