@arcblock/crawler 1.1.6 → 1.3.0

package/lib/cjs/store/index.js

@@ -1,57 +1,49 @@
 "use strict";
-var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
-    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
-    return new (P || (P = Promise))(function (resolve, reject) {
-        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
-        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
-        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
-        step((generator = generator.apply(thisArg, _arguments || [])).next());
-    });
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __exportStar = (this && this.__exportStar) || function(m, exports) {
+    for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
 };
 var __importDefault = (this && this.__importDefault) || function (mod) {
     return (mod && mod.__esModule) ? mod : { "default": mod };
 };
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.initDatabase = initDatabase;
+exports.sequelize = void 0;
 const core_1 = require("@sequelize/core");
 const sqlite3_1 = require("@sequelize/sqlite3");
 const path_1 = __importDefault(require("path"));
 const config_1 = require("../config");
 const job_1 = require("./job");
 const snapshot_1 = require("./snapshot");
-function initDatabase() {
-    return __awaiter(this, void 0, void 0, function* () {
-        const sequelize = new core_1.Sequelize({
-            dialect: sqlite3_1.SqliteDialect,
-            storage: path_1.default.join(config_1.config.dataDir, 'snap-kit.db'),
-            logging: (msg) => process.env.SQLITE_LOG && config_1.logger.debug(msg),
-            pool: {
-                min: 0,
-                max: 10,
-                idle: 10000,
-            },
-            retry: {
-                match: [/SQLITE_BUSY/],
-                name: 'query',
-                max: 10,
-            },
-        });
-        job_1.Job.initModel(sequelize);
-        snapshot_1.Snapshot.initModel(sequelize);
-        try {
-            yield Promise.all([
-                sequelize.query('pragma journal_mode = WAL;'),
-                sequelize.query('pragma synchronous = normal;'),
-                sequelize.query('pragma journal_size_limit = 67108864;'),
-            ]);
-            yield sequelize.authenticate();
-            yield sequelize.sync({ alter: process.env.ALTER_SQLITE === 'true' });
-            config_1.logger.info('Successfully connected to database');
-        }
-        catch (error) {
-            config_1.logger.error('Failed to connect to database:', error);
-            throw error;
-        }
-        return sequelize;
-    });
-}
+const sequelize = new core_1.Sequelize({
+    dialect: sqlite3_1.SqliteDialect,
+    storage: path_1.default.join(config_1.config.dataDir, 'snap-kit.db'),
+    logging: (msg) => process.env.SQLITE_LOG && config_1.logger.debug(msg),
+    pool: {
+        min: 0,
+        max: 10,
+        idle: 10000,
+    },
+    retry: {
+        match: [/SQLITE_BUSY/],
+        name: 'query',
+        max: 10,
+    },
+});
+exports.sequelize = sequelize;
+sequelize.query('pragma journal_mode = WAL;');
+sequelize.query('pragma synchronous = normal;');
+sequelize.query('pragma journal_size_limit = 67108864;');
+job_1.Job.initModel(sequelize);
+snapshot_1.Snapshot.initModel(sequelize);
+__exportStar(require("./job"), exports);
+__exportStar(require("./snapshot"), exports);

package/lib/cjs/store/job.d.ts

@@ -9,9 +9,14 @@ export interface JobState {
     width?: number;
     height?: number;
     quality?: number;
+    format?: 'png' | 'jpeg' | 'webp';
     timeout?: number;
     fullPage?: boolean;
     lastModified?: string;
+    waitTime?: number;
+    replace?: boolean;
+    sync?: boolean;
+    ignoreRobots?: boolean;
     headers?: Record<string, string>;
     cookies?: CookieParam[];
     localStorage?: {

package/lib/cjs/store/migrate.d.ts

@@ -0,0 +1,4 @@
+import { Umzug } from 'umzug';
+declare const umzug: Umzug<import("@sequelize/sqlite3").SqliteQueryInterface<import("@sequelize/sqlite3").SqliteDialect>>;
+export declare function migrate(): Promise<import("umzug").MigrationMeta[]>;
+export { umzug };

package/lib/cjs/store/migrate.js

@@ -0,0 +1,63 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.umzug = void 0;
+exports.migrate = migrate;
+/* eslint-disable global-require */
+const umzug_1 = require("umzug");
+const index_1 = require("./index");
+const migration20250615 = __importStar(require("./migrations/20250615-genesis"));
+const migration20250616Replace = __importStar(require("./migrations/20250616-replace"));
+const umzug = new umzug_1.Umzug({
+    migrations: [
+        {
+            name: '20250615-genesis',
+            up: ({ context }) => migration20250615.up({ context }),
+            down: ({ context }) => migration20250615.down({ context }),
+        },
+        {
+            name: '20250616-replace',
+            up: ({ context }) => migration20250616Replace.up({ context }),
+            down: ({ context }) => migration20250616Replace.down({ context }),
+        },
+    ],
+    context: index_1.sequelize.getQueryInterface(),
+    storage: new umzug_1.SequelizeStorage({ sequelize: index_1.sequelize }),
+    logger: console,
+});
+exports.umzug = umzug;
+function migrate() {
+    return umzug.up();
+}

package/lib/cjs/store/migrations/20250615-genesis.d.ts

@@ -0,0 +1,6 @@
+export declare function up({ context }: {
+    context: any;
+}): Promise<void>;
+export declare function down({ context }: {
+    context: any;
+}): Promise<void>;

package/lib/cjs/store/migrations/20250615-genesis.js

@@ -0,0 +1,114 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.up = up;
+exports.down = down;
+/* eslint-disable no-console */
+const core_1 = require("@sequelize/core");
+function up(_a) {
+    return __awaiter(this, arguments, void 0, function* ({ context }) {
+        console.log('[20250615-genesis:up] Migrating...');
+        yield context.createTable('snap', {
+            jobId: {
+                type: core_1.DataTypes.STRING,
+                primaryKey: true,
+                allowNull: false,
+            },
+            url: {
+                type: core_1.DataTypes.STRING,
+                allowNull: false,
+                index: true,
+            },
+            status: {
+                type: core_1.DataTypes.ENUM('success', 'failed', 'pending'),
+                allowNull: false,
+            },
+            html: {
+                type: core_1.DataTypes.TEXT,
+                allowNull: true,
+            },
+            screenshot: {
+                type: core_1.DataTypes.STRING,
+                allowNull: true,
+            },
+            error: {
+                type: core_1.DataTypes.STRING,
+                allowNull: true,
+            },
+            lastModified: {
+                type: core_1.DataTypes.STRING,
+                allowNull: true,
+            },
+            meta: {
+                type: core_1.DataTypes.JSON,
+                allowNull: true,
+            },
+            options: {
+                type: core_1.DataTypes.JSON,
+                allowNull: true,
+            },
+            createdAt: {
+                type: core_1.DataTypes.DATE,
+                defaultValue: core_1.DataTypes.NOW,
+            },
+            updatedAt: {
+                type: core_1.DataTypes.DATE,
+                defaultValue: core_1.DataTypes.NOW,
+            },
+        });
+        yield context.createTable('jobs', {
+            id: {
+                type: core_1.DataTypes.STRING(40),
+                primaryKey: true,
+            },
+            queue: {
+                type: core_1.DataTypes.STRING(32),
+                allowNull: false,
+            },
+            job: {
+                type: core_1.DataTypes.JSON,
+                allowNull: false,
+            },
+            retryCount: {
+                type: core_1.DataTypes.INTEGER,
+            },
+            delay: {
+                type: core_1.DataTypes.INTEGER,
+            },
+            willRunAt: {
+                type: core_1.DataTypes.INTEGER,
+            },
+            cancelled: {
+                type: core_1.DataTypes.BOOLEAN,
+                defaultValue: false,
+            },
+            createdAt: {
+                type: core_1.DataTypes.DATE,
+                defaultValue: core_1.DataTypes.NOW,
+                index: true,
+            },
+            updatedAt: {
+                type: core_1.DataTypes.DATE,
+                defaultValue: core_1.DataTypes.NOW,
+                index: true,
+            },
+        });
+        console.log('[20250615-genesis:up] Migrated successfully!');
+    });
+}
+function down(_a) {
+    return __awaiter(this, arguments, void 0, function* ({ context }) {
+        console.log('[20250615-genesis:down] Migrating...');
+        yield context.dropTable('snap');
+        yield context.dropTable('jobs');
+        console.log('[20250615-genesis:down] Migrated successfully!');
+    });
+}

package/lib/cjs/store/migrations/20250616-replace.d.ts

@@ -0,0 +1,6 @@
+export declare function up({ context }: {
+    context: any;
+}): Promise<void>;
+export declare function down({ context }: {
+    context: any;
+}): Promise<void>;

package/lib/cjs/store/migrations/20250616-replace.js

@@ -0,0 +1,40 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.up = up;
+exports.down = down;
+/* eslint-disable no-console */
+const core_1 = require("@sequelize/core");
+function up(_a) {
+    return __awaiter(this, arguments, void 0, function* ({ context }) {
+        console.log('[20250616-replace:up] Migrating...');
+        yield context.addColumn('snap', 'replace', {
+            type: core_1.DataTypes.BOOLEAN,
+            allowNull: false,
+            defaultValue: false,
+            index: true,
+        });
+        yield context.addIndex('snap', ['createdAt']);
+        yield context.addIndex('snap', ['updatedAt']);
+        yield context.addIndex('snap', ['status']);
+        console.log('[20250616-replace:up] Migrated successfully!');
+    });
+}
+function down(_a) {
+    return __awaiter(this, arguments, void 0, function* ({ context }) {
+        console.log('[20250616-replace:down] Migrating...');
+        yield context.removeColumn('snap', 'replace');
+        yield context.removeIndex('snap', ['createdAt']);
+        yield context.removeIndex('snap', ['updatedAt']);
+        yield context.removeIndex('snap', ['status']);
+        console.log('[20250616-replace:down] Migrated successfully!');
+    });
+}

package/lib/cjs/store/snapshot.d.ts

@@ -8,6 +8,7 @@ export interface SnapshotModel {
     screenshot?: string | null;
     error?: string;
     lastModified?: string;
+    replace?: boolean;
     meta?: {
         title?: string;
         description?: string;
@@ -35,6 +36,7 @@ export declare class Snapshot extends Model<SnapshotModel> implements SnapshotMo
     screenshot?: SnapshotModel['screenshot'];
     error?: SnapshotModel['error'];
     lastModified?: SnapshotModel['lastModified'];
+    replace?: SnapshotModel['replace'];
     meta?: SnapshotModel['meta'];
     options: SnapshotModel['options'];
     static initModel(sequelize: Sequelize): typeof Snapshot;

package/lib/cjs/store/snapshot.js

@@ -27,6 +27,7 @@ class Snapshot extends core_1.Model {
             status: {
                 type: core_1.DataTypes.ENUM('success', 'failed', 'pending'),
                 allowNull: false,
+                index: true,
             },
             html: {
                 type: core_1.DataTypes.TEXT,
@@ -44,6 +45,12 @@ class Snapshot extends core_1.Model {
                 type: core_1.DataTypes.STRING,
                 allowNull: true,
             },
+            replace: {
+                type: core_1.DataTypes.BOOLEAN,
+                allowNull: false,
+                defaultValue: false,
+                index: true,
+            },
             meta: {
                 type: core_1.DataTypes.JSON,
                 allowNull: true,

package/lib/esm/crawler.d.ts

@@ -1,11 +1,15 @@
-import { JobState } from './store/job';
-import { SnapshotModel } from './store/snapshot';
-export declare function createCrawlQueue(): void;
+import { Page } from '@blocklet/puppeteer';
+import { JobState, SnapshotModel } from './store';
+type PageHandler = {
+    handleScreenshot?: (page: Page, params?: JobState) => Promise<Buffer | null>;
+    handleHtml?: (page: Page, params?: JobState) => Promise<string | null>;
+};
+export declare function createCrawlQueue(queue: string, handler?: PageHandler): any;
 export declare function getDataDir(): Promise<{
     htmlDir: string;
     screenshotDir: string;
 }>;
-export declare const getPageContent: ({ url, includeScreenshot, includeHtml, width, height, quality, timeout, fullPage, headers, cookies, localStorage, }: JobState) => Promise<{
+export declare const getPageContent: ({ url, includeScreenshot, includeHtml, width, height, quality, format, timeout, waitTime, fullPage, headers, cookies, localStorage, }: JobState, handler?: PageHandler) => Promise<{
     html: string | null;
     screenshot: Uint8Array<ArrayBufferLike> | null;
     meta: {
@@ -18,4 +22,7 @@ export declare const getPageContent: ({ url, includeScreenshot, includeHtml, wid
  * @param params
  * @param callback callback when job finished
  */
+export declare function enqueue(queue: any, params: Omit<JobState, 'jobId'>, callback?: (snapshot: SnapshotModel | null) => void): Promise<string>;
 export declare function crawlUrl(params: Omit<JobState, 'jobId'>, callback?: (snapshot: SnapshotModel | null) => void): Promise<string>;
+export declare function crawlCode(params: Omit<JobState, 'jobId'>, callback?: (snapshot: SnapshotModel | null) => void): Promise<string>;
+export {};

package/lib/esm/crawler.js

@@ -14,44 +14,44 @@ import fs from 'fs-extra';
 import path from 'path';
 import { config, logger } from './config';
 import { initPage } from './puppeteer';
-import { convertJobToSnapshot, formatSnapshot } from './services/snapshot';
-import { Job } from './store/job';
-import { Snapshot } from './store/snapshot';
-import { findMaxScrollHeight, formatUrl, isAcceptCrawler, md5 } from './utils';
+import { createCarbonImage } from './services/carbon';
+import { convertJobToSnapshot, deleteSnapshots, formatSnapshot } from './services/snapshot';
+import { Job, Snapshot, sequelize } from './store';
+import { findMaxScrollHeight, formatUrl, isAcceptCrawler, md5, sleep } from './utils';
 const { BaseState } = require('@abtnode/models');
-let crawlQueue;
-export function createCrawlQueue() {
+// eslint-disable-next-line import/no-mutable-exports
+const crawlQueue = createCrawlQueue('urlCrawler');
+const syncQueue = createCrawlQueue('syncCrawler');
+const codeQueue = createCrawlQueue('codeCrawler', {
+    handleScreenshot: createCarbonImage,
+});
+export function createCrawlQueue(queue, handler) {
     const db = new BaseState(Job);
-    crawlQueue = createQueue({
-        store: new SequelizeStore(db, 'crawler'),
+    return createQueue({
+        store: new SequelizeStore(db, queue),
         concurrency: config.concurrency,
         onJob: (job) => __awaiter(this, void 0, void 0, function* () {
             logger.info('Starting to execute crawl job', job);
-            const canCrawl = yield isAcceptCrawler(job.url);
-            if (!canCrawl) {
-                logger.error(`failed to crawl ${job.url}, denied by robots.txt`, job);
-                const snapshot = convertJobToSnapshot({
-                    job,
-                    snapshot: {
-                        status: 'failed',
-                        error: 'Denied by robots.txt',
-                    },
-                });
-                yield Snapshot.upsert(snapshot);
-                return snapshot;
+            // check robots.txt
+            if (!job.ignoreRobots) {
+                const canCrawl = yield isAcceptCrawler(job.url);
+                if (!canCrawl) {
+                    logger.error(`failed to crawl ${job.url}, denied by robots.txt`, job);
+                    const snapshot = convertJobToSnapshot({
+                        job,
+                        snapshot: {
+                            status: 'failed',
+                            error: 'Denied by robots.txt',
+                        },
+                    });
+                    yield Snapshot.upsert(snapshot);
+                    return snapshot;
+                }
             }
-            // if index reach autoCloseBrowserCount, close browser
-            // try {
-            //   if (index >= autoCloseBrowserCount) {
-            //     await closeBrowser({ trimCache: false });
-            //   }
-            // } catch (error) {
-            //   logger.error('failed to close browser when queue index reached autoCloseBrowserCount:', error);
-            // }
             const formattedJob = Object.assign(Object.assign({}, job), { cookies: (config.cookies || []).concat(job.cookies || []), localStorage: (config.localStorage || []).concat(job.localStorage || []), url: formatUrl(job.url) });
             try {
                 // get page content later
-                const result = yield getPageContent(formattedJob);
+                const result = yield getPageContent(formattedJob, handler);
                 if (!result || (!result.html && !result.screenshot)) {
                     logger.error(`failed to crawl ${formattedJob.url}, empty content`, formattedJob);
                     const snapshot = convertJobToSnapshot({
@@ -64,22 +64,40 @@ export function createCrawlQueue() {
                     yield Snapshot.upsert(snapshot);
                     return snapshot;
                 }
-                // save html and screenshot to data dir
-                const { screenshotPath, htmlPath } = yield saveSnapshotToLocal({
-                    screenshot: result.screenshot,
-                    html: result.html,
-                });
-                // const lastModified = job.lastmodMap?.get(url) || new Date().toISOString();
-                const snapshot = convertJobToSnapshot({
-                    job: formattedJob,
-                    snapshot: {
-                        status: 'success',
-                        screenshot: screenshotPath === null || screenshotPath === void 0 ? void 0 : screenshotPath.replace(config.dataDir, ''),
-                        html: htmlPath === null || htmlPath === void 0 ? void 0 : htmlPath.replace(config.dataDir, ''),
-                        meta: result.meta,
-                    },
-                });
-                yield Snapshot.upsert(snapshot);
+                const snapshot = yield sequelize.transaction((txn) => __awaiter(this, void 0, void 0, function* () {
+                    // delete old snapshot
+                    if (formattedJob.replace) {
+                        try {
+                            const deletedJobIds = yield deleteSnapshots({
+                                url: formattedJob.url,
+                                replace: true,
+                            }, { txn });
+                            if (deletedJobIds) {
+                                logger.info('Deleted old snapshot', { deletedJobIds });
+                            }
+                        }
+                        catch (error) {
+                            logger.error('Failed to delete old snapshot', { error, formattedJob });
+                        }
+                    }
+                    // save html and screenshot to data dir
+                    const { screenshotPath, htmlPath } = yield saveSnapshotToLocal({
+                        screenshot: result.screenshot,
+                        html: result.html,
+                        format: formattedJob.format,
+                    });
+                    const snapshot = convertJobToSnapshot({
+                        job: formattedJob,
+                        snapshot: {
+                            status: 'success',
+                            screenshot: screenshotPath === null || screenshotPath === void 0 ? void 0 : screenshotPath.replace(config.dataDir, ''),
+                            html: htmlPath === null || htmlPath === void 0 ? void 0 : htmlPath.replace(config.dataDir, ''),
+                            meta: result.meta,
+                        },
+                    });
+                    yield Snapshot.upsert(snapshot, { transaction: txn });
+                    return snapshot;
+                }));
                 return snapshot;
             }
             catch (error) {
@@ -107,13 +125,13 @@ export function getDataDir() {
     });
 }
 function saveSnapshotToLocal(_a) {
-    return __awaiter(this, arguments, void 0, function* ({ screenshot, html }) {
+    return __awaiter(this, arguments, void 0, function* ({ screenshot, html, format = 'webp', }) {
         const { htmlDir, screenshotDir } = yield getDataDir();
         let screenshotPath = null;
         let htmlPath = null;
         if (screenshot) {
             const hash = md5(screenshot);
-            screenshotPath = path.join(screenshotDir, `${hash}.webp`);
+            screenshotPath = path.join(screenshotDir, `${hash}.${format}`);
             logger.debug('saveSnapshotToLocal.screenshot', { screenshotPath });
             yield fs.writeFile(screenshotPath, screenshot);
         }
@@ -129,7 +147,7 @@ function saveSnapshotToLocal(_a) {
         };
     });
 }
-export const getPageContent = (_a) => __awaiter(void 0, [_a], void 0, function* ({ url, includeScreenshot = true, includeHtml = true, width = 1440, height = 900, quality = 80, timeout = 90 * 1000, fullPage = false, headers, cookies, localStorage, }) {
+export const getPageContent = (_a, handler_1) => __awaiter(void 0, [_a, handler_1], void 0, function* ({ url, includeScreenshot = true, includeHtml = true, width = 1440, height = 900, quality = 80, format = 'webp', timeout = 90 * 1000, waitTime = 0, fullPage = false, headers, cookies, localStorage, }, handler) {
     const page = yield initPage();
     if (width && height) {
         yield page.setViewport({ width, height, deviceScaleFactor: 2 });
@@ -166,9 +184,18 @@ export const getPageContent = (_a) => __awaiter(void 0, [_a], void 0, function*
         }
         // await for networkidle0
         // https://pptr.dev/api/puppeteer.page.waitfornetworkidle
-        yield page.waitForNetworkIdle({
-            idleTime: 1.5 * 1000,
-        });
+        try {
+            yield Promise.all([
+                page.waitForNetworkIdle({
+                    idleTime: 1.5 * 1000,
+                    timeout,
+                }),
+                sleep(waitTime),
+            ]);
+        }
+        catch (err) {
+            logger.warn(`Failed to wait for network idle in ${url}:`, err);
+        }
         // get screenshot
         if (includeScreenshot) {
             // Try to find the tallest element and set the browser to the same height
@@ -184,7 +211,9 @@ export const getPageContent = (_a) => __awaiter(void 0, [_a], void 0, function*
                 }
             }
             try {
-                screenshot = yield page.screenshot({ fullPage, quality, type: 'webp' });
+                screenshot = (handler === null || handler === void 0 ? void 0 : handler.handleScreenshot)
+                    ? yield handler.handleScreenshot(page)
+                    : yield page.screenshot({ fullPage, quality, type: format });
             }
             catch (err) {
                 logger.error('Failed to get screenshot:', err);
@@ -212,12 +241,12 @@ export const getPageContent = (_a) => __awaiter(void 0, [_a], void 0, function*
             // check if the page is an error page
             const isErrorPage = ['<h2>Unexpected Application Error!</h2>', 'Current route occurred an error'].some((errorHtml) => data.html.includes(errorHtml));
             if (isErrorPage) {
-                throw new Error('Page is an error page');
+                throw new Error(`${url} is an error page`);
             }
             meta.title = data.title;
             meta.description = data.description;
             if (includeHtml) {
-                html = data.html;
+                html = (handler === null || handler === void 0 ? void 0 : handler.handleHtml) ? yield handler.handleHtml(page) : data.html;
             }
         }
         catch (err) {
@@ -244,17 +273,17 @@ export const getPageContent = (_a) => __awaiter(void 0, [_a], void 0, function*
  * @param callback callback when job finished
  */
 // eslint-disable-next-line require-await
-export function crawlUrl(params, callback) {
+export function enqueue(queue, params, callback) {
     return __awaiter(this, void 0, void 0, function* () {
         // skip duplicate job
         const existsJob = yield Job.isExists(params);
-        if (existsJob) {
+        if (existsJob && !params.sync) {
             logger.info(`Crawl job already exists for ${params.url}, skip`);
             return existsJob.id;
         }
         logger.info('enqueue crawl job', params);
         const jobId = randomUUID();
-        const job = crawlQueue.push(Object.assign(Object.assign({}, params), { id: jobId }));
+        const job = queue.push(Object.assign(Object.assign({}, params), { id: jobId }));
         job.on('finished', (_a) => __awaiter(this, [_a], void 0, function* ({ result }) {
             logger.info(`Crawl completed ${params.url}, status: ${result ? 'success' : 'failed'}`, { job: params, result });
             callback === null || callback === void 0 ? void 0 : callback(result ? yield formatSnapshot(result) : null);
@@ -266,3 +295,9 @@ export function crawlUrl(params, callback) {
         return jobId;
     });
 }
+export function crawlUrl(params, callback) {
+    return enqueue(params.sync ? syncQueue : crawlQueue, params, callback);
+}
+export function crawlCode(params, callback) {
+    return enqueue(codeQueue, Object.assign({ ignoreRobots: true, includeHtml: false, includeScreenshot: true }, params), callback);
+}