优化爬虫，增加缓存，线程池，保存浏览器用户信息

patches/puppeteer-api/lib/profile-dir.js

@@ -0,0 +1,77 @@
+/**
+ * 按 sessionKey 解析持久 Chrome Profile 目录
+ */
+const fs = require('fs');
+const path = require('path');
+const crypto = require('crypto');
+const { RUNTIME_DIR, PERSIST_BROWSER_PROFILE, PROFILE_MAX_AGE_MS } = require('./constants');
+
+const PROFILES_DIR = path.join(RUNTIME_DIR, 'profiles');
+
+/**
+ * @param {string} sessionKey
+ * @param {'standard'|'real'} profile
+ * @returns {string|null}
+ */
+function resolveProfileDir(sessionKey, profile = 'standard') {
+    if (!sessionKey || !PERSIST_BROWSER_PROFILE) {
+        return null;
+    }
+    const hash = crypto.createHash('md5').update(String(sessionKey)).digest('hex');
+    const dir = path.join(PROFILES_DIR, `${hash}-${profile}`);
+    fs.mkdirSync(dir, { recursive: true, mode: 0o700 });
+    return dir;
+}
+
+function profileExists(sessionKey, profile = 'standard') {
+    const dir = resolveProfileDir(sessionKey, profile);
+    if (!dir) return false;
+    return fs.existsSync(dir);
+}
+
+function pruneStaleProfiles() {
+    if (!PERSIST_BROWSER_PROFILE || PROFILE_MAX_AGE_MS <= 0) {
+        return 0;
+    }
+    if (!fs.existsSync(PROFILES_DIR)) {
+        return 0;
+    }
+    const now = Date.now();
+    let removed = 0;
+    for (const name of fs.readdirSync(PROFILES_DIR)) {
+        const full = path.join(PROFILES_DIR, name);
+        try {
+            const stat = fs.statSync(full);
+            if (!stat.isDirectory()) continue;
+            if (now - stat.mtimeMs > PROFILE_MAX_AGE_MS) {
+                fs.rmSync(full, { recursive: true, force: true });
+                removed++;
+            }
+        } catch (_) {
+            /* ignore */
+        }
+    }
+    return removed;
+}
+
+function getProfileStats() {
+    if (!fs.existsSync(PROFILES_DIR)) {
+        return { profilesDir: PROFILES_DIR, count: 0 };
+    }
+    const count = fs.readdirSync(PROFILES_DIR).filter((name) => {
+        try {
+            return fs.statSync(path.join(PROFILES_DIR, name)).isDirectory();
+        } catch (_) {
+            return false;
+        }
+    }).length;
+    return { profilesDir: PROFILES_DIR, count };
+}
+
+module.exports = {
+    PROFILES_DIR,
+    resolveProfileDir,
+    profileExists,
+    pruneStaleProfiles,
+    getProfileStats,
+};