@cloud411716/fancy-webnovel 0.1.76 → 0.1.78

package/infra.js

@@ -5,13 +5,14 @@
  *   notify(session, text)       — 打印消息给用户（不触发 LLM）
  *   llmFollowup(invocation, prompt) — 发送消息给 LLM，让 LLM 继续处理
  *   spawnScript(cmd, args, opts)    — spawn 子进程，解析 stdout JSON
+ *   startActivity(session)  — 开始 Activity Line 计时，返回 stop 函数
  *
  * 所有 fancy-* 插件均直接 import 此文件。
  */
 
-// ---------------------------------------------------------------------------
+// --------------------------------------------------------------------------
 // notify — 打印消息给用户（queueMicrotask 避免打断 LLM 输出顺序）
-// ---------------------------------------------------------------------------
+// --------------------------------------------------------------------------
 
 export function notify(session, text) {
   queueMicrotask(() => {
@@ -25,7 +26,7 @@ export function notify(session, text) {
 
 // --------------------------------------------------------------------------
 // llmFollowup
-// ---------------------------------------------------------------------------
+// --------------------------------------------------------------------------
 
 export function llmFollowup(invocation, content) {
   const prompt = typeof content === 'string' ? content : JSON.stringify(content);
@@ -35,19 +36,37 @@ export function llmFollowup(invocation, content) {
   });
 }
 
-// ---------------------------------------------------------------------------
+// --------------------------------------------------------------------------
+// startActivity — Activity Line 计时器（显示 Running… + 实时计时）
+// 调用 startActivity() 返回 stop 函数，进入新阶段时先 stop 再 start
+// --------------------------------------------------------------------------
+
+/**
+ * @param {import('@deepseek-ai/dsh-session').Session} session
+ * @returns {() => void} stop — 调用后停止计时
+ */
+export function startActivity(session) {
+  session.append('turn/start', { source: 'fancy-plugin' });
+  return () => session.append('turn/end', { source: 'fancy-plugin' });
+}
+
+// --------------------------------------------------------------------------
 // spawnScript — spawn 子进程，stdout 只收集 JSON，stderr 单独收集
-// ---------------------------------------------------------------------------
+// opts.stop() 会在进程终止时自动调用（用于清理 turn/start）
+// --------------------------------------------------------------------------
 
-export function spawnScript(cmd, args, { timeout = 30000, cwd } = {}) {
+export function spawnScript(cmd, args, { timeout = 30000, cwd, stop } = {}) {
   return new Promise(async (resolve) => {
     const { spawn } = await import('child_process');
     const proc = spawn(cmd, args, { timeout, cwd, stdio: ['ignore', 'pipe', 'pipe'] });
     const outParts = [], errParts = [];
     proc.stdout.on('data', (d) => { outParts.push(d.toString()); });
     proc.stderr.on('data', (d) => { errParts.push(d.toString()); });
+
+    const cleanup = () => { try { stop?.(); } catch (_) {} };
     proc.on('close', (code) => {
       const stdout = outParts.join(''), stderr = errParts.join('');
+      cleanup();
       try {
         const parsed = JSON.parse(stdout);
         resolve({
@@ -63,20 +82,6 @@ export function spawnScript(cmd, args, { timeout = 30000, cwd } = {}) {
         resolve({ ok: false, output: stdout, error: stderr || ('exit ' + code) });
       }
     });
-    proc.on('error', (err) => resolve({ ok: false, output: '', error: err.message }));
+    proc.on('error', (err) => { cleanup(); resolve({ ok: false, output: '', error: err.message }); });
   });
 }
-
-// --------------------------------------------------------------------------
-// activity — Activity Line 计时器（显示 Running… + 实时计时）
-// 调用 startActivity() 返回 stop 函数，进入新阶段时先 stop 再 start
-// --------------------------------------------------------------------------
-
-/**
- * @param {import('@deepseek-ai/dsh-session').Session} session
- * @returns {() => void} stop — 调用后停止计时
- */
-export function startActivity(session) {
-  session.append('turn/start', { source: 'fancy-plugin' });
-  return () => session.append('turn/end', { source: 'fancy-plugin' });
-}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@cloud411716/fancy-webnovel",
-  "version": "0.1.76",
+  "version": "0.1.78",
   "type": "module",
   "main": "index.js",
   "exports": {

package/plugins/fancy-bootstrap/index.js

@@ -80,8 +80,10 @@ export async function apply(ctx) {
 
       notify(session, bootstrap.notify({ type: 'initializing', projectRoot }));
       const stopActivity = startActivity(session);
-      const result = await spawnScript('node', [join(SCRIPTS_DIR, 'init.js'), projectRoot], { timeout: 30000 });
-      stopActivity();
+      const result = await spawnScript('node', [join(SCRIPTS_DIR, 'init.js'), projectRoot], {
+        timeout: 30000,
+        stop: stopActivity,
+      });
 
       if (result.ok) {
         notify(session, bootstrap.notify({ type: 'success' }));

package/plugins/fancy-scan/index.js

@@ -97,9 +97,8 @@ export async function apply(ctx) {
         'node',
         [join(SCAN_SCRIPTS_DIR, 'run-scan.js'),
          '--project-root', projectRoot, '--platform', platform, '--length', length],
-        { timeout: 120000 }
+        { timeout: 120000, stop: stopActivity }
       );
-      stopActivity();
 
       if (!scriptResult.ok) {
         const err = (typeof scriptResult.error === 'object' && scriptResult.error !== null)

package/plugins/fancy-scan/scripts/run-scan.js

@@ -174,49 +174,27 @@ function spawnScraper(platform, length, outDir) {
   return new Promise(async (resolve) => {
     const { spawn } = await import('child_process');
     const args = scraperArgs(platform, length, outDir);
-    const ports = ['9222', '9223'];
-    let portIdx = 0;
-    let dots = 0;
-    const interval = setInterval(() => {
-      dots++;
-      process.stderr.write('.');
-    }, 3000);
-
-    function tryPort(port) {
-      process.stderr.write(`\n  → 端口 ${port} ...`);
-      const scraperScript = join(dirname(fileURLToPath(import.meta.url)), 'scrapers', args.script);
-      const fullArgs = ['--port', port, ...args.extra];
-      const p = spawn('node', [scraperScript, ...fullArgs], {
-        timeout: 120000,
-        stdio: ['ignore', 'pipe', 'inherit'],
-      });
-      let out = '';
-      p.stdout.on('data', d => { out += d.toString(); });
-      p.on('close', (code) => {
-        clearInterval(interval);
-        if (code === 0 && out.trim()) {
-          resolve({ ok: true, output: out, stderr: '' });
-        } else if (portIdx < ports.length - 1) {
-          portIdx++;
-          process.stderr.write(` (失败)\n`);
-          tryPort(ports[portIdx]);
-        } else {
-          process.stderr.write(` (失败)\n`);
-          const errTail = out.trimEnd().slice(-300);
-          resolve({ ok: false, output: out, error: errTail || `exit ${code}` });
-        }
-      });
-      p.on('error', (e) => {
-        clearInterval(interval);
-        if (portIdx < ports.length - 1) {
-          portIdx++;
-          tryPort(ports[portIdx]);
-        } else {
-          resolve({ ok: false, output: '', error: e.message });
-        }
-      });
-    }
-    tryPort(ports[portIdx]);
+    const scraperScript = join(dirname(fileURLToPath(import.meta.url)), 'scrapers', args.script);
+    const fullArgs = [...args.extra];
+
+    process.stderr.write(`\n→ 启动采集脚本...`);
+    const p = spawn('node', [scraperScript, ...fullArgs], {
+      timeout: 300000,
+      stdio: ['ignore', 'pipe', 'inherit'],
+    });
+    let out = '';
+    p.stdout.on('data', d => { out += d.toString(); });
+    p.on('close', (code) => {
+      if (code === 0 && out.trim()) {
+        resolve({ ok: true, output: out, stderr: '' });
+      } else {
+        const errTail = out.trimEnd().slice(-500);
+        resolve({ ok: false, output: out, error: errTail || `exit ${code}` });
+      }
+    });
+    p.on('error', (e) => {
+      resolve({ ok: false, output: '', error: e.message });
+    });
   });
 }
 
@@ -400,7 +378,7 @@ async function main() {
       console.error(`  ❌ 平台 ${platform} 暂不支持采集`);
       process.exit(1);
     }
-    console.error(`→ 采集 ${PLATFORM_CN[platform]}（Chrome CDP）...`);
+    console.error(`→ 采集 ${PLATFORM_CN[platform]}...`);
     const result = await spawnScraper(platform, length, scanDir);
     if (!result.ok) {
       console.error(`  ❌ ${PLATFORM_CN[platform]} 采集失败：${result.error}`);

package/plugins/fancy-scan/scripts/scraper-registry.md

@@ -14,11 +14,9 @@
 
 ## Chrome CDP 环境要求
 
-> ⚠️ **除 qidian 外，其他所有 scraper 都需要先启动 Chrome CDP**：
-> ```bash
-> node {SKILL_DIR}/browser-cdp/scripts/setup-cdp-chrome.js --yes
-> ```
-> qidian 默认走 mobile-SSR（不需要 Chrome），失败才降级 CDP。
+> ⚠️ **所有 scraper 现在使用 playwright-core 自己管理浏览器，不再依赖 agent-browser 和 Chrome CDP 端口。**
+> 每个 scraper 独立 launch/close 浏览器进程，采集完成后自动关闭。
+> 如需在采集前留时间登录，可传入 `--login-wait <秒>` 参数。
 
 ---
 

package/plugins/fancy-scan/scripts/scrapers/cdp-utils.cjs

@@ -1,11 +1,13 @@
 /**
- * CDP 工具函数 — 各平台采集脚本的公共依赖
+ * 工具函数 — 各平台采集脚本的公共依赖
  *
- * 使用方式：
- *   const { ab, sleep, evalJSON, evalJSONBase64, scrollLoad, getArg, safeStr, localDateStamp } = require("./cdp-utils");
+ * 提供：
+ *   getArg()   — 命令行参数解析
+ *   localDateStamp() — 本地日期戳（YYYYMMDD）
+ *   runCli()   — CLI 入口包装（含进程超时和信号处理）
+ *   sleep()    — 同步阻塞等待（Atomics.wait）
  *
- * 前置：
- *   node {SKILL_DIR}/browser-cdp/scripts/setup-cdp-chrome.js 9222
+ * 注意：cdp-utils 不再依赖 agent-browser/CDP 端口。浏览器生命周期由各 scraper 自己管理。
  */
 
 const { execFileSync } = require("child_process");