init

apps/backend/src/services/ocr.js

@@ -0,0 +1,213 @@
+import { spawn } from 'node:child_process'
+import readline from 'node:readline'
+import process from 'node:process'
+
+import { runtimeConfig } from '../config/runtime.js'
+
+const DEFAULT_TIMEOUT_MS = 60_000
+
+let workerPromise = null
+let requestSequence = 0
+const pendingRequests = new Map()
+let stderrBuffer = []
+
+export async function recognizeTencentCaptcha(payload) {
+  return callOcrWorker('recognize', payload)
+}
+
+export async function batchRecognizeTencentCaptcha(payload) {
+  return callOcrWorker('batch', payload, { timeoutMs: 180_000 })
+}
+
+export async function warmupLocalOcrWorker() {
+  await ensureOcrWorker()
+}
+
+export async function closeLocalOcrWorker() {
+  const worker = await workerPromise?.catch(() => null)
+
+  if (!worker) {
+    workerPromise = null
+    return
+  }
+
+  worker.child.kill()
+  workerPromise = null
+}
+
+async function callOcrWorker(action, payload, { timeoutMs = DEFAULT_TIMEOUT_MS } = {}) {
+  const worker = await ensureOcrWorker()
+  const requestId = `ocr-${Date.now()}-${++requestSequence}`
+
+  return await new Promise((resolve, reject) => {
+    const timer = setTimeout(() => {
+      pendingRequests.delete(requestId)
+      worker.child.kill()
+      reject(new Error('本地 OCR 识别超时，请确认 uv 环境和 OCR worker 依赖正常'))
+    }, timeoutMs)
+
+    pendingRequests.set(requestId, {
+      resolve: (value) => {
+        clearTimeout(timer)
+        resolve(value)
+      },
+      reject: (error) => {
+        clearTimeout(timer)
+        reject(error)
+      },
+    })
+
+    worker.child.stdin.write(
+      `${JSON.stringify({ id: requestId, action, payload: payload || {} })}\n`,
+      'utf8',
+    )
+  })
+}
+
+async function ensureOcrWorker() {
+  if (!workerPromise) {
+    workerPromise = createOcrWorker().catch((error) => {
+      workerPromise = null
+      throw error
+    })
+  }
+
+  return workerPromise
+}
+
+function createOcrWorker() {
+  return new Promise((resolve, reject) => {
+    const projectRoot = resolveOcrProjectRoot()
+    stderrBuffer = []
+
+    const child = spawn('uv', ['run', 'ocr-worker', 'worker'], {
+      cwd: projectRoot,
+      stdio: ['pipe', 'pipe', 'pipe'],
+      env: process.env,
+    })
+    child.stdin.setDefaultEncoding('utf8')
+
+    const stdoutReader = readline.createInterface({ input: child.stdout })
+    const worker = { child, stdoutReader }
+
+    const readyTimer = setTimeout(() => {
+      reject(new Error(`本地 OCR worker 启动超时，请检查 ${projectRoot} 下是否已执行 uv sync`))
+      child.kill()
+    }, 15_000)
+
+    let ready = false
+
+    stdoutReader.on('line', (line) => {
+      const message = tryParseJson(line)
+
+      if (!message || typeof message !== 'object') {
+        return
+      }
+
+      if (message.type === 'ready') {
+        if (!ready) {
+          ready = true
+          clearTimeout(readyTimer)
+          resolve(worker)
+        }
+        return
+      }
+
+      if (message.type !== 'response') {
+        return
+      }
+
+      const requestId = String(message.id || '')
+      const pending = pendingRequests.get(requestId)
+
+      if (!pending) {
+        return
+      }
+
+      pendingRequests.delete(requestId)
+      pending.resolve(message.payload)
+    })
+
+    child.stderr.on('data', (chunk) => {
+      const text = String(chunk || '').trim()
+
+      if (!text) {
+        return
+      }
+
+      stderrBuffer.push(text)
+      if (stderrBuffer.length > 20) {
+        stderrBuffer = stderrBuffer.slice(-20)
+      }
+    })
+
+    child.on('error', (error) => {
+      clearTimeout(readyTimer)
+      const workerError = buildWorkerError(error)
+      failAllPendingRequests(workerError)
+      if (!ready) {
+        reject(workerError)
+      }
+      workerPromise = null
+    })
+
+    child.on('exit', (code, signal) => {
+      clearTimeout(readyTimer)
+      const workerError = new Error(
+        buildWorkerExitMessage({
+          code,
+          signal,
+          projectRoot,
+          stderrText: stderrBuffer.join('\n'),
+        }),
+      )
+      failAllPendingRequests(workerError)
+      if (!ready) {
+        reject(workerError)
+      }
+      workerPromise = null
+    })
+  })
+}
+
+function failAllPendingRequests(error) {
+  for (const [requestId, pending] of pendingRequests.entries()) {
+    pendingRequests.delete(requestId)
+    pending.reject(error)
+  }
+}
+
+function resolveOcrProjectRoot() {
+  return String(runtimeConfig.ocr.projectRoot || '').trim()
+}
+
+function buildWorkerError(error) {
+  if (error instanceof Error && error.message.includes('spawn uv ENOENT')) {
+    return new Error('未找到 uv 命令，请先安装 uv，并确认它在 PATH 里')
+  }
+
+  return new Error(
+    `本地 OCR worker 启动失败: ${error instanceof Error ? error.message : String(error || '未知错误')}`,
+  )
+}
+
+function buildWorkerExitMessage({ code, signal, projectRoot, stderrText }) {
+  const reason = signal
+    ? `signal ${signal}`
+    : typeof code === 'number'
+      ? `exit code ${code}`
+      : 'unknown reason'
+
+  const installHint = `请先执行: cd ${projectRoot} && uv sync`
+  const stderrHint = stderrText ? `\n${stderrText}` : ''
+
+  return `本地 OCR worker 已退出（${reason}）。${installHint}${stderrHint}`
+}
+
+function tryParseJson(text) {
+  try {
+    return JSON.parse(text)
+  } catch {
+    return null
+  }
+}