Function parseWeFlow

packages/parser/src/formats/weflow.ts:156–605

(options: ParseOptions)

// ==================== 解析器实现 ====================

async function* parseWeFlow(options: ParseOptions): AsyncGenerator<ParseEvent, void, unknown> {
  const { filePath, batchSize = 5000, onProgress, onLog } = options

  const totalBytes = getFileSize(filePath)
  let bytesRead = 0
  let messagesProcessed = 0

  // 发送初始进度
  const initialProgress = createProgress('parsing', 0, totalBytes, 0, '')
  yield { type: 'progress', data: initialProgress }
  onProgress?.(initialProgress)

  // 记录解析开始
  onLog?.('info', `开始解析 WeFlow 导出文件，大小: ${(totalBytes / 1024 / 1024).toFixed(2)} MB`)

  // 读取文件头获取基本信息
  const headContent = readFileHeadBytes(filePath, 5000)

  // 使用流式读取获取完整的 session 对象（因为 session.avatar 可能很大）
  let session: WeFlowSession | null = null
  try {
    await new Promise<void>((resolve) => {
      const sessionStream = fs.createReadStream(filePath, { encoding: 'utf-8' })

      let sessionContent = ''
      let inSession = false
      let braceDepth = 0
      let inString = false
      let escape = false

      sessionStream.on('data', (chunk: string | Buffer) => {
        const str = typeof chunk === 'string' ? chunk : chunk.toString()

        for (let i = 0; i < str.length; i++) {
          const char = str[i]

          if (!inSession) {
            // 查找 "session": 的位置
            const searchStr = '"session":'
            if (str.slice(i, i + searchStr.length) === searchStr) {
              inSession = true
              i += searchStr.length - 1
              continue
            }
          } else {
            sessionContent += char

            if (escape) {
              escape = false
              continue
            }

            if (char === '\\' && inString) {
              escape = true
              continue
            }

            if (char === '"') {