@crawlee/jsdom 4.0.0-beta.14 → 4.0.0-beta.141

package/README.md

@@ -1,23 +1,23 @@
 <h1 align="center">
     <a href="https://crawlee.dev">
         <picture>
-          <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/apify/crawlee/master/website/static/img/crawlee-dark.svg?sanitize=true">
-          <img alt="Crawlee" src="https://raw.githubusercontent.com/apify/crawlee/master/website/static/img/crawlee-light.svg?sanitize=true" width="500">
+          <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/apify/crawlee/master/website/static/img/crawlee-dark.svg?sanitize=true" />
+          <img alt="Crawlee" src="https://raw.githubusercontent.com/apify/crawlee/master/website/static/img/crawlee-light.svg?sanitize=true" width="500" />
         </picture>
     </a>
-    <br>
+    <br />
     <small>A web scraping and browser automation library</small>
 </h1>
 
-<p align=center>
-    <a href="https://trendshift.io/repositories/5179" target="_blank"><img src="https://trendshift.io/api/badge/repositories/5179" alt="apify%2Fcrawlee | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+<p align="center">
+    <a href="https://trendshift.io/repositories/5179" target="_blank"><img src="https://trendshift.io/api/badge/repositories/5179" alt="apify%2Fcrawlee | Trendshift" width="250" height="55"/></a>
 </p>
 
-<p align=center>
-    <a href="https://www.npmjs.com/package/@crawlee/core" rel="nofollow"><img src="https://img.shields.io/npm/v/@crawlee/core.svg" alt="NPM latest version" data-canonical-src="https://img.shields.io/npm/v/@crawlee/core/next.svg" style="max-width: 100%;"></a>
-    <a href="https://www.npmjs.com/package/@crawlee/core" rel="nofollow"><img src="https://img.shields.io/npm/dm/@crawlee/core.svg" alt="Downloads" data-canonical-src="https://img.shields.io/npm/dm/@crawlee/core.svg" style="max-width: 100%;"></a>
-    <a href="https://discord.gg/jyEM2PRvMU" rel="nofollow"><img src="https://img.shields.io/discord/801163717915574323?label=discord" alt="Chat on discord" data-canonical-src="https://img.shields.io/discord/801163717915574323?label=discord" style="max-width: 100%;"></a>
-    <a href="https://github.com/apify/crawlee/actions/workflows/test-ci.yml"><img src="https://github.com/apify/crawlee/actions/workflows/test-ci.yml/badge.svg?branch=master" alt="Build Status" style="max-width: 100%;"></a>
+<p align="center">
+    <a href="https://www.npmjs.com/package/@crawlee/core" rel="nofollow"><img src="https://img.shields.io/npm/v/@crawlee/core.svg" alt="NPM latest version" data-canonical-src="https://img.shields.io/npm/v/@crawlee/core/next.svg" /></a>
+    <a href="https://www.npmjs.com/package/@crawlee/core" rel="nofollow"><img src="https://img.shields.io/npm/dm/@crawlee/core.svg" alt="Downloads" data-canonical-src="https://img.shields.io/npm/dm/@crawlee/core.svg" /></a>
+    <a href="https://discord.gg/jyEM2PRvMU" rel="nofollow"><img src="https://img.shields.io/discord/801163717915574323?label=discord" alt="Chat on discord" data-canonical-src="https://img.shields.io/discord/801163717915574323?label=discord" /></a>
+    <a href="https://github.com/apify/crawlee/actions/workflows/test-ci.yml"><img src="https://github.com/apify/crawlee/actions/workflows/test-ci.yml/badge.svg?branch=master" alt="Build Status" /></a>
 </p>
 
 Crawlee covers your crawling and scraping end-to-end and **helps you build reliable scrapers. Fast.**
@@ -89,7 +89,7 @@ By default, Crawlee stores data to `./storage` in the current working directory.
 We provide automated beta builds for every merged code change in Crawlee. You can find them in the npm [list of releases](https://www.npmjs.com/package/crawlee?activeTab=versions). If you want to test new features or bug fixes before we release them, feel free to install a beta build like this:
 
 ```bash
-npm install crawlee@3.12.3-beta.13
+npm install crawlee@next
 ```
 
 If you also use the [Apify SDK](https://github.com/apify/apify-sdk-js), you need to specify dependency overrides in your `package.json` file so that you don't end up with multiple versions of Crawlee installed:
@@ -98,9 +98,9 @@ If you also use the [Apify SDK](https://github.com/apify/apify-sdk-js), you need
 {
     "overrides": {
        "apify": {
-           "@crawlee/core": "3.12.3-beta.13",
-           "@crawlee/types": "3.12.3-beta.13",
-           "@crawlee/utils": "3.12.3-beta.13"
+           "@crawlee/core": "$crawlee",
+           "@crawlee/types": "$crawlee",
+           "@crawlee/utils": "$crawlee"
        }
     }
 }

package/index.d.ts

@@ -1,3 +1,2 @@
 export * from '@crawlee/http';
 export * from './internals/jsdom-crawler.js';
-//# sourceMappingURL=index.d.ts.map

package/index.js

@@ -1,3 +1,2 @@
 export * from '@crawlee/http';
 export * from './internals/jsdom-crawler.js';
-//# sourceMappingURL=index.js.map

package/internals/jsdom-crawler.d.ts

@@ -1,13 +1,16 @@
-import type { BasicCrawlingContext, Configuration, EnqueueLinksOptions, ErrorHandler, GetUserDataFromRequest, HttpCrawlerOptions, InternalHttpCrawlingContext, InternalHttpHook, RequestHandler, RequestProvider, RouterRoutes, SkippedRequestCallback } from '@crawlee/http';
+import type { AddRequestsBatchedResult, ContextPipeline, CrawlingContext, EnqueueLinksOptions, ErrorHandler, ExtractLinksOptions, GetUserDataFromRequest, HttpCrawlerOptions, InternalHttpCrawlingContext, InternalHttpHook, RequestHandler, RouterHandler, RouterRoutes, RouteSchemas, RoutesFromSchemas } from '@crawlee/http';
 import { HttpCrawler } from '@crawlee/http';
 import type { Dictionary } from '@crawlee/types';
-import { type CheerioRoot, type RobotsTxtFile } from '@crawlee/utils';
+import type { CheerioAPI } from 'cheerio';
 import type { DOMWindow } from 'jsdom';
 import { VirtualConsole } from 'jsdom';
+import { z } from 'zod';
 export type JSDOMErrorHandler<UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
-JSONData extends Dictionary = any> = ErrorHandler<JSDOMCrawlingContext<UserData, JSONData>>;
-export interface JSDOMCrawlerOptions<ExtendedContext extends JSDOMCrawlingContext = JSDOMCrawlingContext, UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
-JSONData extends Dictionary = any> extends HttpCrawlerOptions<JSDOMCrawlingContext<UserData, JSONData>, ExtendedContext> {
+JSONData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
+ContextExtension = Dictionary<never>> = ErrorHandler<CrawlingContext, JSDOMCrawlingContext<UserData, JSONData> & ContextExtension>;
+export interface JSDOMCrawlerOptions<ContextExtension = Dictionary<never>, ExtendedContext extends JSDOMCrawlingContext = JSDOMCrawlingContext & ContextExtension, UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
+JSONData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
+Routes extends Record<keyof Routes, Dictionary> = Record<string, UserData>, StatisticStateExtension extends object = {}> extends HttpCrawlerOptions<JSDOMCrawlingContext<UserData, JSONData>, ContextExtension, ExtendedContext, Routes, StatisticStateExtension> {
     /**
      * Download and run scripts.
      */
@@ -50,101 +53,140 @@ JSONData extends Dictionary = any> extends InternalHttpCrawlingContext<UserData,
      * });
      * ```
      */
-    parseWithCheerio(selector?: string, timeoutMs?: number): Promise<CheerioRoot>;
+    parseWithCheerio(selector?: string, timeoutMs?: number): Promise<CheerioAPI>;
+    /**
+     * Extracts URLs from the parsed DOM, without adding them to the request queue.
+     */
+    extractLinks(options?: ExtractLinksOptions): Promise<string[]>;
+    /**
+     * Helper function for extracting URLs from the parsed DOM and adding them to the request queue.
+     */
+    enqueueLinks(options?: EnqueueLinksOptions): Promise<AddRequestsBatchedResult>;
 }
 export type JSDOMRequestHandler<UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
 JSONData extends Dictionary = any> = RequestHandler<JSDOMCrawlingContext<UserData, JSONData>>;
-export declare class JSDOMCrawler<ExtendedContext extends JSDOMCrawlingContext = JSDOMCrawlingContext> extends HttpCrawler<JSDOMCrawlingContext, ExtendedContext> {
+export declare class JSDOMCrawler<ContextExtension = Dictionary<never>, ExtendedContext extends JSDOMCrawlingContext = JSDOMCrawlingContext & ContextExtension, Routes extends Record<keyof Routes, Dictionary> = Record<string, GetUserDataFromRequest<JSDOMCrawlingContext['request']>>, StatisticStateExtension extends object = {}> extends HttpCrawler<JSDOMCrawlingContext, ContextExtension, ExtendedContext, Routes, StatisticStateExtension> {
+    #private;
+    /**
+     * @internal
+     */
     protected static optionsShape: {
-// @ts-ignore optional peer dependency or compatibility with es2022
-        runScripts: import("ow").BooleanPredicate & import("ow").BasePredicate<boolean | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        hideInternalConsole: import("ow").BooleanPredicate & import("ow").BasePredicate<boolean | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        navigationTimeoutSecs: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        ignoreSslErrors: import("ow").BooleanPredicate & import("ow").BasePredicate<boolean | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        additionalMimeTypes: import("ow").ArrayPredicate<string>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        suggestResponseEncoding: import("ow").StringPredicate & import("ow").BasePredicate<string | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        forceResponseEncoding: import("ow").StringPredicate & import("ow").BasePredicate<string | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        persistCookiesPerSession: import("ow").BooleanPredicate & import("ow").BasePredicate<boolean | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        additionalHttpErrorStatusCodes: import("ow").ArrayPredicate<number>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        ignoreHttpErrorStatusCodes: import("ow").ArrayPredicate<number>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        preNavigationHooks: import("ow").ArrayPredicate<unknown> & import("ow").BasePredicate<unknown[] | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        postNavigationHooks: import("ow").ArrayPredicate<unknown> & import("ow").BasePredicate<unknown[] | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        contextPipelineBuilder: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        extendContext: import("ow").Predicate<Function> & import("ow").BasePredicate<Function | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        requestList: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        requestQueue: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        requestHandler: import("ow").Predicate<Function> & import("ow").BasePredicate<Function | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        requestHandlerTimeoutSecs: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        errorHandler: import("ow").Predicate<Function> & import("ow").BasePredicate<Function | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        failedRequestHandler: import("ow").Predicate<Function> & import("ow").BasePredicate<Function | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        maxRequestRetries: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        sameDomainDelaySecs: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        maxSessionRotations: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        maxRequestsPerCrawl: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        maxCrawlDepth: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        autoscaledPoolOptions: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        sessionPoolOptions: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        useSessionPool: import("ow").BooleanPredicate & import("ow").BasePredicate<boolean | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        proxyConfiguration: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        statusMessageLoggingInterval: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        statusMessageCallback: import("ow").Predicate<Function> & import("ow").BasePredicate<Function | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        retryOnBlocked: import("ow").BooleanPredicate & import("ow").BasePredicate<boolean | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        respectRobotsTxtFile: import("ow").AnyPredicate<boolean | object>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        onSkippedRequest: import("ow").Predicate<Function> & import("ow").BasePredicate<Function | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        httpClient: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        minConcurrency: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        maxConcurrency: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        maxRequestsPerMinute: import("ow").NumberPredicate & import("ow").BasePredicate<number | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        keepAlive: import("ow").BooleanPredicate & import("ow").BasePredicate<boolean | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        log: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        experiments: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        statisticsOptions: import("ow").ObjectPredicate<object> & import("ow").BasePredicate<object | undefined>;
+        runScripts: z.ZodOptional<z.ZodBoolean>;
+        hideInternalConsole: z.ZodOptional<z.ZodBoolean>;
+        navigationTimeoutSecs: z.ZodDefault<z.ZodCustom<number, number>>;
+        ignoreTlsErrors: z.ZodDefault<z.ZodBoolean>;
+        additionalMimeTypes: z.ZodDefault<z.ZodArray<z.ZodString>>;
+        suggestResponseEncoding: z.ZodOptional<z.ZodString>;
+        forceResponseEncoding: z.ZodOptional<z.ZodString>;
+        saveResponseCookies: z.ZodDefault<z.ZodBoolean>;
+        preNavigationHooks: z.ZodDefault<z.ZodCustom<unknown[], unknown[]>>;
+        postNavigationHooks: z.ZodDefault<z.ZodCustom<unknown[], unknown[]>>;
+        contextPipelineBuilder: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
+        extendContext: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        requestList: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        requestQueue: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        requestManager: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        requestHandler: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        requestHandlerTimeoutSecs: z.ZodOptional<z.ZodCustom<number, number>>;
+        errorHandler: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        failedRequestHandler: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        maxRequestRetries: z.ZodDefault<z.ZodCustom<number, number>>;
+        sameDomainDelaySecs: z.ZodDefault<z.ZodCustom<number, number>>;
+        maxRequestsPerCrawl: z.ZodOptional<z.ZodCustom<number, number>>;
+        maxCrawlDepth: z.ZodOptional<z.ZodCustom<number, number>>;
+        taskLoopOptions: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
+        concurrencySystem: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
+        sessionPool: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        proxyConfiguration: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        statusMessageLoggingInterval: z.ZodDefault<z.ZodCustom<number, number>>;
+        statusMessageCallback: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        additionalHttpErrorStatusCodes: z.ZodDefault<z.ZodArray<z.ZodCustom<number, number>>>;
+        ignoreHttpErrorStatusCodes: z.ZodDefault<z.ZodArray<z.ZodCustom<number, number>>>;
+        blockedStatusCodes: z.ZodOptional<z.ZodArray<z.ZodCustom<number, number>>>;
+        retryOnBlocked: z.ZodDefault<z.ZodBoolean>;
+        respectRobotsTxtFile: z.ZodDefault<z.ZodUnion<readonly [z.ZodBoolean, z.ZodCustom<Dictionary, Dictionary>]>>;
+        transactionalStorage: z.ZodOptional<z.ZodUnion<readonly [z.ZodBoolean, z.ZodObject<{
+            requestQueue: z.ZodOptional<z.ZodEnum<{
+                deferred: "deferred";
+                writeThrough: "writeThrough";
+            }>>;
+        }, z.core.$strict>]>>;
+        onSkippedRequest: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+// @ts-ignore optional peer dependency or compatibility with es2022
+        httpClient: z.ZodOptional<z.ZodCustom<import("@crawlee/http-client").BaseHttpClient, import("@crawlee/http-client").BaseHttpClient>>;
+// @ts-ignore optional peer dependency or compatibility with es2022
+        configuration: z.ZodOptional<z.ZodCustom<import("@crawlee/http").Configuration, import("@crawlee/http").Configuration>>;
+        storageBackend: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+// @ts-ignore optional peer dependency or compatibility with es2022
+        eventManager: z.ZodOptional<z.ZodCustom<import("@crawlee/http").EventManager, import("@crawlee/http").EventManager>>;
+        logger: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        minConcurrency: z.ZodOptional<z.ZodCustom<number, number>>;
+        maxConcurrency: z.ZodOptional<z.ZodCustom<number, number>>;
+        maxRequestsPerMinute: z.ZodOptional<z.ZodCustom<number, number>>;
+        keepAlive: z.ZodOptional<z.ZodBoolean>;
+        statistics: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
+        id: z.ZodOptional<z.ZodString>;
     };
-    protected runScripts: boolean;
-    protected hideInternalConsole: boolean;
-    protected virtualConsole: VirtualConsole | null;
-    constructor(options?: JSDOMCrawlerOptions<ExtendedContext>, config?: Configuration);
+    /** @internal */
+    protected static optionsSchema: z.ZodObject<{
+        runScripts: z.ZodOptional<z.ZodBoolean>;
+        hideInternalConsole: z.ZodOptional<z.ZodBoolean>;
+        navigationTimeoutSecs: z.ZodDefault<z.ZodCustom<number, number>>;
+        ignoreTlsErrors: z.ZodDefault<z.ZodBoolean>;
+        additionalMimeTypes: z.ZodDefault<z.ZodArray<z.ZodString>>;
+        suggestResponseEncoding: z.ZodOptional<z.ZodString>;
+        forceResponseEncoding: z.ZodOptional<z.ZodString>;
+        saveResponseCookies: z.ZodDefault<z.ZodBoolean>;
+        preNavigationHooks: z.ZodDefault<z.ZodCustom<unknown[], unknown[]>>;
+        postNavigationHooks: z.ZodDefault<z.ZodCustom<unknown[], unknown[]>>;
+        contextPipelineBuilder: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
+        extendContext: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        requestList: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        requestQueue: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        requestManager: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        requestHandler: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        requestHandlerTimeoutSecs: z.ZodOptional<z.ZodCustom<number, number>>;
+        errorHandler: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        failedRequestHandler: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        maxRequestRetries: z.ZodDefault<z.ZodCustom<number, number>>;
+        sameDomainDelaySecs: z.ZodDefault<z.ZodCustom<number, number>>;
+        maxRequestsPerCrawl: z.ZodOptional<z.ZodCustom<number, number>>;
+        maxCrawlDepth: z.ZodOptional<z.ZodCustom<number, number>>;
+        taskLoopOptions: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
+        concurrencySystem: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
+        sessionPool: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        proxyConfiguration: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        statusMessageLoggingInterval: z.ZodDefault<z.ZodCustom<number, number>>;
+        statusMessageCallback: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+        additionalHttpErrorStatusCodes: z.ZodDefault<z.ZodArray<z.ZodCustom<number, number>>>;
+        ignoreHttpErrorStatusCodes: z.ZodDefault<z.ZodArray<z.ZodCustom<number, number>>>;
+        blockedStatusCodes: z.ZodOptional<z.ZodArray<z.ZodCustom<number, number>>>;
+        retryOnBlocked: z.ZodDefault<z.ZodBoolean>;
+        respectRobotsTxtFile: z.ZodDefault<z.ZodUnion<readonly [z.ZodBoolean, z.ZodCustom<Dictionary, Dictionary>]>>;
+        transactionalStorage: z.ZodOptional<z.ZodUnion<readonly [z.ZodBoolean, z.ZodObject<{
+            requestQueue: z.ZodOptional<z.ZodEnum<{
+                deferred: "deferred";
+                writeThrough: "writeThrough";
+            }>>;
+        }, z.core.$strict>]>>;
+        onSkippedRequest: z.ZodOptional<z.ZodCustom<(...args: any[]) => unknown, (...args: any[]) => unknown>>;
+// @ts-ignore optional peer dependency or compatibility with es2022
+        httpClient: z.ZodOptional<z.ZodCustom<import("@crawlee/http-client").BaseHttpClient, import("@crawlee/http-client").BaseHttpClient>>;
+// @ts-ignore optional peer dependency or compatibility with es2022
+        configuration: z.ZodOptional<z.ZodCustom<import("@crawlee/http").Configuration, import("@crawlee/http").Configuration>>;
+        storageBackend: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+// @ts-ignore optional peer dependency or compatibility with es2022
+        eventManager: z.ZodOptional<z.ZodCustom<import("@crawlee/http").EventManager, import("@crawlee/http").EventManager>>;
+        logger: z.ZodOptional<z.ZodType<Dictionary<any>, unknown, z.core.$ZodTypeInternals<Dictionary<any>, unknown>>>;
+        minConcurrency: z.ZodOptional<z.ZodCustom<number, number>>;
+        maxConcurrency: z.ZodOptional<z.ZodCustom<number, number>>;
+        maxRequestsPerMinute: z.ZodOptional<z.ZodCustom<number, number>>;
+        keepAlive: z.ZodOptional<z.ZodBoolean>;
+        statistics: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
+        id: z.ZodOptional<z.ZodString>;
+    }, z.core.$strict>;
+    constructor(options?: JSDOMCrawlerOptions<ContextExtension, ExtendedContext, any, any, Routes, StatisticStateExtension>);
+    protected buildContextPipeline(): ContextPipeline<CrawlingContext, JSDOMCrawlingContext>;
     /**
      * Returns the currently used `VirtualConsole` instance. Can be used to listen for the JSDOM's internal console messages.
      *
@@ -164,24 +206,6 @@ export declare class JSDOMCrawler<ExtendedContext extends JSDOMCrawlingContext =
     private parseContent;
     private addHelpers;
 }
-interface EnqueueLinksInternalOptions {
-    options?: EnqueueLinksOptions;
-    window: DOMWindow | null;
-    requestQueue: RequestProvider;
-    robotsTxtFile?: RobotsTxtFile;
-    onSkippedRequest?: SkippedRequestCallback;
-    originalRequestUrl: string;
-    finalRequestUrl?: string;
-}
-interface BoundEnqueueLinksInternalOptions {
-    enqueueLinks: BasicCrawlingContext['enqueueLinks'];
-    options?: EnqueueLinksOptions;
-    window: DOMWindow | null;
-    originalRequestUrl: string;
-    finalRequestUrl?: string;
-}
-/** @internal */
-export declare function domCrawlerEnqueueLinks(options: EnqueueLinksInternalOptions | BoundEnqueueLinksInternalOptions): Promise<unknown>;
 /**
  * Creates new {@link Router} instance that works based on request labels.
  * This instance can then serve as a `requestHandler` of your {@link JSDOMCrawler}.
@@ -206,7 +230,6 @@ export declare function domCrawlerEnqueueLinks(options: EnqueueLinksInternalOpti
  * await crawler.run();
  * ```
  */
-// @ts-ignore optional peer dependency or compatibility with es2022
-export declare function createJSDOMRouter<Context extends JSDOMCrawlingContext = JSDOMCrawlingContext, UserData extends Dictionary = GetUserDataFromRequest<Context['request']>>(routes?: RouterRoutes<Context, UserData>): import("@crawlee/http").RouterHandler<Context>;
-export {};
-//# sourceMappingURL=jsdom-crawler.d.ts.map
+export declare function createJSDOMRouter<Context extends JSDOMCrawlingContext = JSDOMCrawlingContext, Routes extends Record<keyof Routes, Dictionary> = Record<string, GetUserDataFromRequest<Context['request']>>>(routes?: RouterRoutes<Context, Routes>): RouterHandler<Context, Routes>;
+export declare function createJSDOMRouter<Context extends JSDOMCrawlingContext = JSDOMCrawlingContext, UserData extends Dictionary = GetUserDataFromRequest<Context['request']>>(routes?: RouterRoutes<Context, Record<string, UserData>>): RouterHandler<Context, Record<string, UserData>>;
+export declare function createJSDOMRouter<Context extends JSDOMCrawlingContext = JSDOMCrawlingContext, const Schemas extends RouteSchemas = RouteSchemas>(schemas: Schemas): RouterHandler<Context, RoutesFromSchemas<Schemas>>;

package/internals/jsdom-crawler.js

@@ -1,8 +1,8 @@
-import { enqueueLinks, HttpCrawler, resolveBaseUrlForEnqueueLinksFiltering, Router, tryAbsoluteURL, } from '@crawlee/http';
+import { EnqueueStrategy, HttpCrawler, NavigationSkippedError, resolveBaseUrlForEnqueueLinksFiltering, Router, } from '@crawlee/http';
 import { sleep } from '@crawlee/utils';
-import * as cheerio from 'cheerio';
+import { parseArgument, tryAbsoluteURL } from '@crawlee/utils/internal';
 import { JSDOM, ResourceLoader, VirtualConsole } from 'jsdom';
-import ow from 'ow';
+import { z } from 'zod';
 import { addTimeoutToPromise } from '@apify/timeout';
 /**
  * Provides a framework for the parallel crawling of web pages using plain HTTP requests and
@@ -26,38 +26,40 @@ import { addTimeoutToPromise } from '@apify/timeout';
  * and then invokes the user-provided {@link JSDOMCrawlerOptions.requestHandler} to extract page data
  * using the `window` object.
  *
- * The source URLs are represented using {@link Request} objects that are fed from
- * {@link RequestList} or {@link RequestQueue} instances provided by the {@link JSDOMCrawlerOptions.requestList}
- * or {@link JSDOMCrawlerOptions.requestQueue} constructor options, respectively.
+ * The source URLs are represented using {@link Request} objects that are fed from the
+ * {@link IRequestManager|request manager} provided via the {@link JSDOMCrawlerOptions.requestManager|`requestManager`}
+ * constructor option (a {@link RequestQueue} is itself a request manager). To read from a read-only source such
+ * as a {@link RequestList} while still being able to enqueue new requests, combine it with a queue into a
+ * {@link RequestManagerTandem} via {@link IRequestLoader.toTandem|`requestLoader.toTandem()`} and pass the
+ * result as `requestManager`.
  *
- * If both {@link JSDOMCrawlerOptions.requestList} and {@link JSDOMCrawlerOptions.requestQueue} are used,
- * the instance first processes URLs from the {@link RequestList} and automatically enqueues all of them
- * to {@link RequestQueue} before it starts their processing. This ensures that a single URL is not crawled multiple times.
+ * > The {@link JSDOMCrawlerOptions.requestList|`requestList`} and {@link JSDOMCrawlerOptions.requestQueue|`requestQueue`}
+ * > options are deprecated; they are still accepted and folded into a single `requestManager` for back-compat.
  *
  * The crawler finishes when there are no more {@link Request} objects to crawl.
  *
- * We can use the `preNavigationHooks` to adjust `gotOptions`:
+ * We can use the `preNavigationHooks` to adjust the crawling context before the request is made:
  *
  * ```
  * preNavigationHooks: [
- *     (crawlingContext, gotOptions) => {
+ *     (crawlingContext) => {
  *         // ...
  *     },
  * ]
  * ```
  *
- * By default, `JSDOMCrawler` only processes web pages with the `text/html`
- * and `application/xhtml+xml` MIME content types (as reported by the `Content-Type` HTTP header),
+ * By default, `JSDOMCrawler` only processes web pages with the `text/html`, `application/xhtml+xml`, `text/xml`, `application/xml`,
+ * and `application/json` MIME content types (as reported by the `Content-Type` HTTP header),
  * and skips pages with other content types. If you want the crawler to process other content types,
  * use the {@link JSDOMCrawlerOptions.additionalMimeTypes} constructor option.
  * Beware that the parsing behavior differs for HTML, XML, JSON and other types of content.
  * For more details, see {@link JSDOMCrawlerOptions.requestHandler}.
  *
- * New requests are only dispatched when there is enough free CPU and memory available,
- * using the functionality provided by the {@link AutoscaledPool} class.
- * All {@link AutoscaledPool} configuration options can be passed to the `autoscaledPoolOptions`
- * parameter of the `CheerioCrawler` constructor. For user convenience, the `minConcurrency` and `maxConcurrency`
- * {@link AutoscaledPool} options are available directly in the `CheerioCrawler` constructor.
+ * New requests are only dispatched when there is enough free CPU and memory available, as judged by the crawler's
+ * {@link ConcurrencySystem}.
+ * Concurrency is tuned via the `minConcurrency`, `maxConcurrency` and `maxRequestsPerMinute` options of the
+ * `JSDOMCrawler` constructor, or, for finer control, by injecting a pre-configured
+ * {@link ConcurrencySystem|`concurrencySystem`}.
  *
  * **Example usage:**
  *
@@ -83,30 +85,39 @@ const resources = new ResourceLoader({
     userAgent: 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/107.0.0.0 Safari/537.36',
 });
 export class JSDOMCrawler extends HttpCrawler {
+    /**
+     * @internal
+     */
     static optionsShape = {
         ...HttpCrawler.optionsShape,
-        runScripts: ow.optional.boolean,
-        hideInternalConsole: ow.optional.boolean,
+        runScripts: z.boolean().optional(),
+        hideInternalConsole: z.boolean().optional(),
     };
-    runScripts;
-    hideInternalConsole;
-    virtualConsole = null;
-    constructor(options = {}, config) {
-        const { runScripts = false, hideInternalConsole = false, ...httpOptions } = options;
+    /** @internal */
+    static optionsSchema = z.strictObject(JSDOMCrawler.optionsShape);
+    #runScripts;
+    #hideInternalConsole;
+    #virtualConsole = null;
+    constructor(options = {}) {
+        const { runScripts = false, hideInternalConsole = false, contextPipelineBuilder, ...httpOptions } = parseArgument(options, JSDOMCrawler.optionsSchema, 'JSDOMCrawlerOptions');
         super({
             ...httpOptions,
-            contextPipelineBuilder: () => this.buildContextPipeline()
-                .compose({
-                action: async (context) => await this.parseContent(context),
-                cleanup: async (context) => {
-                    this.getVirtualConsole().off('jsdomError', this.jsdomErrorHandler);
-                    context.window?.close();
-                },
-            })
-                .compose({ action: async (context) => await this.addHelpers(context) }),
-        }, config);
-        this.runScripts = runScripts;
-        this.hideInternalConsole = hideInternalConsole;
+            contextPipelineBuilder: contextPipelineBuilder ?? (() => this.buildContextPipeline()),
+        });
+        this.#runScripts = runScripts;
+        this.#hideInternalConsole = hideInternalConsole;
+    }
+    buildContextPipeline() {
+        return super
+            .buildContextPipeline()
+            .compose({
+            action: async (context) => await this.parseContent(context),
+            cleanup: async (context) => {
+                this.getVirtualConsole().off('jsdomError', this.jsdomErrorHandler);
+                context.window?.close();
+            },
+        })
+            .compose({ action: async (context) => await this.addHelpers(context) });
     }
     /**
      * Returns the currently used `VirtualConsole` instance. Can be used to listen for the JSDOM's internal console messages.
@@ -123,86 +134,116 @@ export class JSDOMCrawler extends HttpCrawler {
      * ```
      */
     getVirtualConsole() {
-        if (this.virtualConsole) {
-            return this.virtualConsole;
+        if (this.#virtualConsole) {
+            return this.#virtualConsole;
         }
-        this.virtualConsole = new VirtualConsole();
-        if (!this.hideInternalConsole) {
-            this.virtualConsole.sendTo(console, { omitJSDOMErrors: true });
+        this.#virtualConsole = new VirtualConsole();
+        if (!this.#hideInternalConsole) {
+            this.#virtualConsole.sendTo(console, { omitJSDOMErrors: true });
         }
-        this.virtualConsole.on('jsdomError', this.jsdomErrorHandler);
-        return this.virtualConsole;
+        this.#virtualConsole.on('jsdomError', this.jsdomErrorHandler);
+        return this.#virtualConsole;
     }
-    jsdomErrorHandler = (error) => this.log.debug('JSDOM error from console', error);
+    jsdomErrorHandler = (error) => this.log.debug('JSDOM error from console', { error });
     async parseContent(crawlingContext) {
-        const isXml = crawlingContext.contentType.type.includes('xml');
-        // TODO handle non-string
-        const { window } = new JSDOM(crawlingContext.body.toString(), {
-            url: crawlingContext.response.url,
-            contentType: isXml ? 'text/xml' : 'text/html',
-            runScripts: this.runScripts ? 'dangerously' : undefined,
-            resources,
-            virtualConsole: this.getVirtualConsole(),
-            pretendToBeVisual: true,
-        });
-        // add some stubs in place of missing API so processing won't fail
-        Object.defineProperty(window, 'matchMedia', {
-            writable: true,
-            value: (query) => ({
-                matches: false,
-                media: query,
-                onchange: null,
-                addListener: () => { },
-                removeListener: () => { },
-                addEventListener: () => { },
-                removeEventListener: () => { },
-                dispatchEvent: () => { },
-            }),
-        });
-        window.document.createRange = () => {
-            const range = new window.Range();
-            range.getBoundingClientRect = () => ({});
-            range.getClientRects = () => ({ item: () => null, length: 0 });
-            return range;
-        };
-        if (this.runScripts) {
-            try {
-                await addTimeoutToPromise(async () => {
-                    return new Promise((resolve) => {
-                        window.addEventListener('load', () => {
-                            resolve();
-                        }, false);
-                    }).catch();
-                }, 10_000, 'Window.load event not fired after 10 seconds.').catch();
+        try {
+            const isXml = crawlingContext.contentType.type.includes('xml');
+            // TODO handle non-string
+            const { window } = new JSDOM(crawlingContext.body.toString(), {
+                url: crawlingContext.response.url,
+                contentType: isXml ? 'text/xml' : 'text/html',
+                runScripts: this.#runScripts ? 'dangerously' : undefined,
+                resources,
+                virtualConsole: this.getVirtualConsole(),
+                pretendToBeVisual: true,
+            });
+            // add some stubs in place of missing API so processing won't fail
+            Object.defineProperty(window, 'matchMedia', {
+                writable: true,
+                value: (query) => ({
+                    matches: false,
+                    media: query,
+                    onchange: null,
+                    addListener: () => { },
+                    removeListener: () => { },
+                    addEventListener: () => { },
+                    removeEventListener: () => { },
+                    dispatchEvent: () => { },
+                }),
+            });
+            window.document.createRange = () => {
+                const range = new window.Range();
+                range.getBoundingClientRect = () => ({});
+                range.getClientRects = () => ({ item: () => null, length: 0 });
+                return range;
+            };
+            if (this.#runScripts) {
+                try {
+                    await addTimeoutToPromise(async () => {
+                        return new Promise((resolve) => {
+                            window.addEventListener('load', () => {
+                                resolve();
+                            }, false);
+                        }).catch();
+                    }, 10_000, 'Window.load event not fired after 10 seconds.').catch();
+                }
+                catch (e) {
+                    this.log.debug(e.message);
+                }
             }
-            catch (e) {
-                this.log.debug(e.message);
+            return {
+                window,
+                get body() {
+                    return window.document.documentElement.outerHTML;
+                },
+                get document() {
+                    return window.document;
+                },
+            };
+        }
+        catch (err) {
+            if (err instanceof NavigationSkippedError) {
+                return {
+                    get window() {
+                        throw new NavigationSkippedError('The `window` property is not available - `skipNavigation` was used', { cause: err });
+                    },
+                    get body() {
+                        throw new NavigationSkippedError('The `body` property is not available - `skipNavigation` was used', { cause: err });
+                    },
+                    get document() {
+                        throw new NavigationSkippedError('The `document` property is not available - `skipNavigation` was used', { cause: err });
+                    },
+                };
             }
+            throw err;
         }
-        return {
-            window,
-            get body() {
-                return window.document.documentElement.outerHTML;
-            },
-            get document() {
-                return window.document;
-            },
-        };
     }
     async addHelpers(crawlingContext) {
+        const addRequests = crawlingContext.addRequests;
+        const extractLinks = async (options) => {
+            if (!crawlingContext.window) {
+                throw new Error('Cannot extract links because the JSDOM is not available.');
+            }
+            return extractUrlsFromWindow(crawlingContext.window, options?.selector ?? 'a', options?.baseUrl ?? crawlingContext.request.loadedUrl ?? crawlingContext.request.url);
+        };
         return {
-            enqueueLinks: async (enqueueOptions) => {
-                return domCrawlerEnqueueLinks({
-                    options: { ...enqueueOptions, limit: this.calculateEnqueuedRequestLimit(enqueueOptions?.limit) },
-                    window: crawlingContext.window,
-                    requestQueue: await this.getRequestQueue(),
-                    robotsTxtFile: await this.getRobotsTxtFileForUrl(crawlingContext.request.url),
-                    onSkippedRequest: this.handleSkippedRequest,
-                    originalRequestUrl: crawlingContext.request.url,
+            extractLinks,
+            enqueueLinks: async (options = {}) => {
+                const baseUrl = resolveBaseUrlForEnqueueLinksFiltering({
+                    enqueueStrategy: options.strategy,
                     finalRequestUrl: crawlingContext.request.loadedUrl,
+                    originalRequestUrl: crawlingContext.request.url,
+                    userProvidedBaseUrl: options.baseUrl,
+                });
+                const urls = await extractLinks(options);
+                return addRequests(urls, {
+                    ...options,
+                    baseUrl,
+                    strategy: options.strategy ?? EnqueueStrategy.SameHostname,
                 });
             },
             async waitForSelector(selector, timeoutMs = 5_000) {
+                const cheerio = await import('cheerio');
                 const $ = cheerio.load(crawlingContext.body);
                 if ($(selector).get().length === 0) {
                     if (timeoutMs) {
@@ -214,6 +255,7 @@ export class JSDOMCrawler extends HttpCrawler {
                 }
             },
             async parseWithCheerio(selector, _timeoutMs = 5_000) {
+                const cheerio = await import('cheerio');
                 const $ = cheerio.load(crawlingContext.body);
                 if (selector && $(selector).get().length === 0) {
                     throw new Error(`Selector '${selector}' not found.`);
@@ -223,39 +265,6 @@ export class JSDOMCrawler extends HttpCrawler {
         };
     }
 }
-/** @internal */
-function containsEnqueueLinks(options) {
-    return !!options.enqueueLinks;
-}
-/** @internal */
-export async function domCrawlerEnqueueLinks(options) {
-    const { options: enqueueLinksOptions, window, originalRequestUrl, finalRequestUrl } = options;
-    if (!window) {
-        throw new Error('Cannot enqueue links because the JSDOM is not available.');
-    }
-    const baseUrl = resolveBaseUrlForEnqueueLinksFiltering({
-        enqueueStrategy: enqueueLinksOptions?.strategy,
-        finalRequestUrl,
-        originalRequestUrl,
-        userProvidedBaseUrl: enqueueLinksOptions?.baseUrl,
-    });
-    const urls = extractUrlsFromWindow(window, enqueueLinksOptions?.selector ?? 'a', enqueueLinksOptions?.baseUrl ?? finalRequestUrl ?? originalRequestUrl);
-    if (containsEnqueueLinks(options)) {
-        return options.enqueueLinks({
-            urls,
-            baseUrl,
-            ...enqueueLinksOptions,
-        });
-    }
-    return enqueueLinks({
-        requestQueue: options.requestQueue,
-        robotsTxtFile: options.robotsTxtFile,
-        onSkippedRequest: options.onSkippedRequest,
-        urls,
-        baseUrl,
-        ...enqueueLinksOptions,
-    });
-}
 /**
  * Extracts URLs from a given Window object.
  * @ignore
@@ -272,31 +281,6 @@ function extractUrlsFromWindow(window, selector, baseUrl) {
     })
         .filter((href) => href !== undefined && href !== '');
 }
-/**
- * Creates new {@link Router} instance that works based on request labels.
- * This instance can then serve as a `requestHandler` of your {@link JSDOMCrawler}.
- * Defaults to the {@link JSDOMCrawlingContext}.
- *
- * > Serves as a shortcut for using `Router.create<JSDOMCrawlingContext>()`.
- *
- * ```ts
- * import { JSDOMCrawler, createJSDOMRouter } from 'crawlee';
- *
- * const router = createJSDOMRouter();
- * router.addHandler('label-a', async (ctx) => {
- *    ctx.log.info('...');
- * });
- * router.addDefaultHandler(async (ctx) => {
- *    ctx.log.info('...');
- * });
- *
- * const crawler = new JSDOMCrawler({
- *     requestHandler: router,
- * });
- * await crawler.run();
- * ```
- */
-export function createJSDOMRouter(routes) {
-    return Router.create(routes);
+export function createJSDOMRouter(routesOrSchemas) {
+    return Router.create(routesOrSchemas);
 }
-//# sourceMappingURL=jsdom-crawler.js.map

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@crawlee/jsdom",
-    "version": "4.0.0-beta.14",
+    "version": "4.0.0-beta.141",
     "description": "The scalable web crawling and scraping library for JavaScript/Node.js. Enables development of data extraction and web automation jobs (not only) with headless Chrome and Puppeteer.",
     "engines": {
         "node": ">=22.0.0"
@@ -38,7 +38,7 @@
     },
     "homepage": "https://crawlee.dev",
     "scripts": {
-        "build": "yarn clean && yarn compile && yarn copy",
+        "build": "pnpm clean && pnpm compile && pnpm copy",
         "clean": "rimraf ./dist",
         "compile": "tsc -p tsconfig.build.json",
         "copy": "tsx ../../scripts/copy.ts"
@@ -47,16 +47,16 @@
         "access": "public"
     },
     "dependencies": {
-        "@apify/timeout": "^0.3.0",
+        "@apify/timeout": "^0.4.4",
         "@apify/utilities": "^2.7.10",
-        "@crawlee/http": "4.0.0-beta.14",
-        "@crawlee/types": "4.0.0-beta.14",
-        "@crawlee/utils": "4.0.0-beta.14",
+        "@crawlee/http": "4.0.0-beta.141",
+        "@crawlee/types": "4.0.0-beta.141",
+        "@crawlee/utils": "4.0.0-beta.141",
         "@types/jsdom": "^21.1.7",
         "cheerio": "^1.0.0",
         "jsdom": "^26.1.0",
-        "ow": "^2.0.0",
-        "tslib": "^2.8.1"
+        "tslib": "^2.8.1",
+        "zod": "^4.4.3"
     },
     "lerna": {
         "command": {
@@ -65,5 +65,5 @@
             }
         }
     },
-    "gitHead": "3d889c957289f962078c1e095f33897595b42632"
+    "gitHead": "6f97ccc5beda6b4ec0455b26323dba311a5799e4"
 }

package/index.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,cAAc,eAAe,CAAC;AAC9B,cAAc,8BAA8B,CAAC"}

package/index.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,cAAc,eAAe,CAAC;AAC9B,cAAc,8BAA8B,CAAC"}

package/internals/jsdom-crawler.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"jsdom-crawler.d.ts","sourceRoot":"","sources":["../../src/internals/jsdom-crawler.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACR,oBAAoB,EACpB,aAAa,EACb,mBAAmB,EACnB,YAAY,EACZ,sBAAsB,EACtB,kBAAkB,EAClB,2BAA2B,EAC3B,gBAAgB,EAChB,cAAc,EACd,eAAe,EACf,YAAY,EACZ,sBAAsB,EACzB,MAAM,eAAe,CAAC;AACvB,OAAO,EAEH,WAAW,EAId,MAAM,eAAe,CAAC;AACvB,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,gBAAgB,CAAC;AACjD,OAAO,EAAE,KAAK,WAAW,EAAE,KAAK,aAAa,EAAS,MAAM,gBAAgB,CAAC;AAE7E,OAAO,KAAK,EAAE,SAAS,EAAE,MAAM,OAAO,CAAC;AACvC,OAAO,EAAyB,cAAc,EAAE,MAAM,OAAO,CAAC;AAK9D,MAAM,MAAM,iBAAiB,CACzB,QAAQ,SAAS,UAAU,GAAG,GAAG,EAAE,2EAA2E;AAC9G,QAAQ,SAAS,UAAU,GAAG,GAAG,IACjC,YAAY,CAAC,oBAAoB,CAAC,QAAQ,EAAE,QAAQ,CAAC,CAAC,CAAC;AAE3D,MAAM,WAAW,mBAAmB,CAChC,eAAe,SAAS,oBAAoB,GAAG,oBAAoB,EACnE,QAAQ,SAAS,UAAU,GAAG,GAAG,EAAE,2EAA2E;AAC9G,QAAQ,SAAS,UAAU,GAAG,GAAG,CACnC,SAAQ,kBAAkB,CAAC,oBAAoB,CAAC,QAAQ,EAAE,QAAQ,CAAC,EAAE,eAAe,CAAC;IACnF;;OAEG;IACH,UAAU,CAAC,EAAE,OAAO,CAAC;IACrB;;OAEG;IACH,mBAAmB,CAAC,EAAE,OAAO,CAAC;CACjC;AAED,MAAM,MAAM,SAAS,CACjB,QAAQ,SAAS,UAAU,GAAG,GAAG,EAAE,2EAA2E;AAC9G,QAAQ,SAAS,UAAU,GAAG,GAAG,IACjC,gBAAgB,CAAC,oBAAoB,CAAC,QAAQ,EAAE,QAAQ,CAAC,CAAC,CAAC;AAE/D,MAAM,WAAW,oBAAoB,CACjC,QAAQ,SAAS,UAAU,GAAG,GAAG,EAAE,2EAA2E;AAC9G,QAAQ,SAAS,UAAU,GAAG,GAAG,CACnC,SAAQ,2BAA2B,CAAC,QAAQ,EAAE,QAAQ,CAAC;IACrD,MAAM,EAAE,SAAS,CAAC;IAClB,QAAQ,EAAE,QAAQ,CAAC;IAEnB,IAAI,EAAE,MAAM,CAAC;IAEb;;;;;;;;;;;;OAYG;IACH,eAAe,CAAC,QAAQ,EAAE,MAAM,EAAE,SAAS,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC;IAErE;;;;;;;;;;;OAWG;IACH,gBAAgB,CAAC,QAAQ,CAAC,EAAE,MAAM,EAAE,SAAS,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC,WAAW,CAAC,CAAC;CACjF;AAED,MAAM,MAAM,mBAAmB,CAC3B,QAAQ,SAAS,UAAU,GAAG,GAAG,EAAE,2EAA2E;AAC9G,QAAQ,SAAS,UAAU,GAAG,GAAG,IACjC,cAAc,CAAC,oBAAoB,CAAC,QAAQ,EAAE,QAAQ,CAAC,CAAC,CAAC;AAkF7D,qBAAa,YAAY,CAAC,eAAe,SAAS,oBAAoB,GAAG,oBAAoB,CAAE,SAAQ,WAAW,CAC9G,oBAAoB,EACpB,eAAe,CAClB;IACG,iBAA0B,YAAY;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;MAIpC;IAEF,SAAS,CAAC,UAAU,EAAE,OAAO,CAAC;IAC9B,SAAS,CAAC,mBAAmB,EAAE,OAAO,CAAC;IACvC,SAAS,CAAC,cAAc,EAAE,cAAc,GAAG,IAAI,CAAQ;gBAE3C,OAAO,GAAE,mBAAmB,CAAC,eAAe,CAAM,EAAE,MAAM,CAAC,EAAE,aAAa;IAwBtF;;;;;;;;;;;;;OAaG;IACH,iBAAiB;IAgBjB,OAAO,CAAC,QAAQ,CAAC,iBAAiB,CAAuE;YAE3F,YAAY;YAmEZ,UAAU;CAqC3B;AAED,UAAU,2BAA2B;IACjC,OAAO,CAAC,EAAE,mBAAmB,CAAC;IAC9B,MAAM,EAAE,SAAS,GAAG,IAAI,CAAC;IACzB,YAAY,EAAE,eAAe,CAAC;IAC9B,aAAa,CAAC,EAAE,aAAa,CAAC;IAC9B,gBAAgB,CAAC,EAAE,sBAAsB,CAAC;IAC1C,kBAAkB,EAAE,MAAM,CAAC;IAC3B,eAAe,CAAC,EAAE,MAAM,CAAC;CAC5B;AAED,UAAU,gCAAgC;IACtC,YAAY,EAAE,oBAAoB,CAAC,cAAc,CAAC,CAAC;IACnD,OAAO,CAAC,EAAE,mBAAmB,CAAC;IAC9B,MAAM,EAAE,SAAS,GAAG,IAAI,CAAC;IACzB,kBAAkB,EAAE,MAAM,CAAC;IAC3B,eAAe,CAAC,EAAE,MAAM,CAAC;CAC5B;AASD,gBAAgB;AAChB,wBAAsB,sBAAsB,CAAC,OAAO,EAAE,2BAA2B,GAAG,gCAAgC,oBAoCnH;AAmBD;;;;;;;;;;;;;;;;;;;;;;;GAuBG;AACH,wBAAgB,iBAAiB,CAC7B,OAAO,SAAS,oBAAoB,GAAG,oBAAoB,EAC3D,QAAQ,SAAS,UAAU,GAAG,sBAAsB,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,EAC1E,MAAM,CAAC,EAAE,YAAY,CAAC,OAAO,EAAE,QAAQ,CAAC,kDAEzC"}

package/internals/jsdom-crawler.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"jsdom-crawler.js","sourceRoot":"","sources":["../../src/internals/jsdom-crawler.ts"],"names":[],"mappings":"AAcA,OAAO,EACH,YAAY,EACZ,WAAW,EACX,sCAAsC,EACtC,MAAM,EACN,cAAc,GACjB,MAAM,eAAe,CAAC;AAEvB,OAAO,EAAwC,KAAK,EAAE,MAAM,gBAAgB,CAAC;AAC7E,OAAO,KAAK,OAAO,MAAM,SAAS,CAAC;AAEnC,OAAO,EAAE,KAAK,EAAE,cAAc,EAAE,cAAc,EAAE,MAAM,OAAO,CAAC;AAC9D,OAAO,EAAE,MAAM,IAAI,CAAC;AAEpB,OAAO,EAAE,mBAAmB,EAAE,MAAM,gBAAgB,CAAC;AAuErD;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAwEG;AACH,MAAM,SAAS,GAAG,IAAI,cAAc,CAAC;IACjC,6EAA6E;IAC7E,kDAAkD;IAClD,SAAS,EACL,uHAAuH;CAC9H,CAAC,CAAC;AAEH,MAAM,OAAO,YAAkF,SAAQ,WAGtG;IACa,MAAM,CAAU,YAAY,GAAG;QACrC,GAAG,WAAW,CAAC,YAAY;QAC3B,UAAU,EAAE,EAAE,CAAC,QAAQ,CAAC,OAAO;QAC/B,mBAAmB,EAAE,EAAE,CAAC,QAAQ,CAAC,OAAO;KAC3C,CAAC;IAEQ,UAAU,CAAU;IACpB,mBAAmB,CAAU;IAC7B,cAAc,GAA0B,IAAI,CAAC;IAEvD,YAAY,UAAgD,EAAE,EAAE,MAAsB;QAClF,MAAM,EAAE,UAAU,GAAG,KAAK,EAAE,mBAAmB,GAAG,KAAK,EAAE,GAAG,WAAW,EAAE,GAAG,OAAO,CAAC;QAEpF,KAAK,CACD;YACI,GAAG,WAAW;YACd,sBAAsB,EAAE,GAAG,EAAE,CACzB,IAAI,CAAC,oBAAoB,EAAE;iBACtB,OAAO,CAAC;gBACL,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,EAAE,CAAC,MAAM,IAAI,CAAC,YAAY,CAAC,OAAO,CAAC;gBAC3D,OAAO,EAAE,KAAK,EAAE,OAAO,EAAE,EAAE;oBACvB,IAAI,CAAC,iBAAiB,EAAE,CAAC,GAAG,CAAC,YAAY,EAAE,IAAI,CAAC,iBAAiB,CAAC,CAAC;oBACnE,OAAO,CAAC,MAAM,EAAE,KAAK,EAAE,CAAC;gBAC5B,CAAC;aACJ,CAAC;iBACD,OAAO,CAAC,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,EAAE,CAAC,MAAM,IAAI,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;SAClF,EACD,MAAM,CACT,CAAC;QAEF,IAAI,CAAC,UAAU,GAAG,UAAU,CAAC;QAC7B,IAAI,CAAC,mBAAmB,GAAG,mBAAmB,CAAC;IACnD,CAAC;IAED;;;;;;;;;;;;;OAaG;IACH,iBAAiB;QACb,IAAI,IAAI,CAAC,cAAc,EAAE,CAAC;YACtB,OAAO,IAAI,CAAC,cAAc,CAAC;QAC/B,CAAC;QAED,IAAI,CAAC,cAAc,GAAG,IAAI,cAAc,EAAE,CAAC;QAE3C,IAAI,CAAC,IAAI,CAAC,mBAAmB,EAAE,CAAC;YAC5B,IAAI,CAAC,cAAc,CAAC,MAAM,CAAC,OAAO,EAAE,EAAE,eAAe,EAAE,IAAI,EAAE,CAAC,CAAC;QACnE,CAAC;QAED,IAAI,CAAC,cAAc,CAAC,EAAE,CAAC,YAAY,EAAE,IAAI,CAAC,iBAAiB,CAAC,CAAC;QAE7D,OAAO,IAAI,CAAC,cAAc,CAAC;IAC/B,CAAC;IAEgB,iBAAiB,GAAG,CAAC,KAAY,EAAE,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,0BAA0B,EAAE,KAAK,CAAC,CAAC;IAEjG,KAAK,CAAC,YAAY,CAAC,eAA4C;QACnE,MAAM,KAAK,GAAG,eAAe,CAAC,WAAW,CAAC,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC;QAE/D,yBAAyB;QACzB,MAAM,EAAE,MAAM,EAAE,GAAG,IAAI,KAAK,CAAC,eAAe,CAAC,IAAI,CAAC,QAAQ,EAAE,EAAE;YAC1D,GAAG,EAAE,eAAe,CAAC,QAAQ,CAAC,GAAG;YACjC,WAAW,EAAE,KAAK,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,WAAW;YAC7C,UAAU,EAAE,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,aAAa,CAAC,CAAC,CAAC,SAAS;YACvD,SAAS;YACT,cAAc,EAAE,IAAI,CAAC,iBAAiB,EAAE;YACxC,iBAAiB,EAAE,IAAI;SAC1B,CAAC,CAAC;QAEH,kEAAkE;QAClE,MAAM,CAAC,cAAc,CAAC,MAAM,EAAE,YAAY,EAAE;YACxC,QAAQ,EAAE,IAAI;YACd,KAAK,EAAE,CAAC,KAAc,EAAO,EAAE,CAAC,CAAC;gBAC7B,OAAO,EAAE,KAAK;gBACd,KAAK,EAAE,KAAK;gBACZ,QAAQ,EAAE,IAAI;gBACd,WAAW,EAAE,GAAG,EAAE,GAAE,CAAC;gBACrB,cAAc,EAAE,GAAG,EAAE,GAAE,CAAC;gBACxB,gBAAgB,EAAE,GAAG,EAAE,GAAE,CAAC;gBAC1B,mBAAmB,EAAE,GAAG,EAAE,GAAE,CAAC;gBAC7B,aAAa,EAAE,GAAG,EAAE,GAAE,CAAC;aAC1B,CAAC;SACL,CAAC,CAAC;QACH,MAAM,CAAC,QAAQ,CAAC,WAAW,GAAG,GAAG,EAAE;YAC/B,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,KAAK,EAAE,CAAC;YACjC,KAAK,CAAC,qBAAqB,GAAG,GAAG,EAAE,CAAC,CAAC,EAAE,CAAQ,CAAC;YAChD,KAAK,CAAC,cAAc,GAAG,GAAG,EAAE,CAAC,CAAC,EAAE,IAAI,EAAE,GAAG,EAAE,CAAC,IAAW,EAAE,MAAM,EAAE,CAAC,EAAE,CAAQ,CAAC;YAC7E,OAAO,KAAK,CAAC;QACjB,CAAC,CAAC;QAEF,IAAI,IAAI,CAAC,UAAU,EAAE,CAAC;YAClB,IAAI,CAAC;gBACD,MAAM,mBAAmB,CACrB,KAAK,IAAI,EAAE;oBACP,OAAO,IAAI,OAAO,CAAO,CAAC,OAAO,EAAE,EAAE;wBACjC,MAAM,CAAC,gBAAgB,CACnB,MAAM,EACN,GAAG,EAAE;4BACD,OAAO,EAAE,CAAC;wBACd,CAAC,EACD,KAAK,CACR,CAAC;oBACN,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;gBACf,CAAC,EACD,MAAM,EACN,+CAA+C,CAClD,CAAC,KAAK,EAAE,CAAC;YACd,CAAC;YAAC,OAAO,CAAC,EAAE,CAAC;gBACT,IAAI,CAAC,GAAG,CAAC,KAAK,CAAE,CAAW,CAAC,OAAO,CAAC,CAAC;YACzC,CAAC;QACL,CAAC;QAED,OAAO;YACH,MAAM;YACN,IAAI,IAAI;gBACJ,OAAO,MAAM,CAAC,QAAQ,CAAC,eAAe,CAAC,SAAS,CAAC;YACrD,CAAC;YACD,IAAI,QAAQ;gBACR,OAAO,MAAM,CAAC,QAAQ,CAAC;YAC3B,CAAC;SACJ,CAAC;IACN,CAAC;IAEO,KAAK,CAAC,UAAU,CAAC,eAAkF;QACvG,OAAO;YACH,YAAY,EAAE,KAAK,EAAE,cAAoC,EAAE,EAAE;gBACzD,OAAO,sBAAsB,CAAC;oBAC1B,OAAO,EAAE,EAAE,GAAG,cAAc,EAAE,KAAK,EAAE,IAAI,CAAC,6BAA6B,CAAC,cAAc,EAAE,KAAK,CAAC,EAAE;oBAChG,MAAM,EAAE,eAAe,CAAC,MAAM;oBAC9B,YAAY,EAAE,MAAM,IAAI,CAAC,eAAe,EAAE;oBAC1C,aAAa,EAAE,MAAM,IAAI,CAAC,sBAAsB,CAAC,eAAe,CAAC,OAAO,CAAC,GAAG,CAAC;oBAC7E,gBAAgB,EAAE,IAAI,CAAC,oBAAoB;oBAC3C,kBAAkB,EAAE,eAAe,CAAC,OAAO,CAAC,GAAG;oBAC/C,eAAe,EAAE,eAAe,CAAC,OAAO,CAAC,SAAS;iBACrD,CAAC,CAAC;YACP,CAAC;YACD,KAAK,CAAC,eAAe,CAAC,QAAgB,EAAE,SAAS,GAAG,KAAK;gBACrD,MAAM,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,eAAe,CAAC,IAAI,CAAC,CAAC;gBAE7C,IAAI,CAAC,CAAC,QAAQ,CAAC,CAAC,GAAG,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;oBACjC,IAAI,SAAS,EAAE,CAAC;wBACZ,MAAM,KAAK,CAAC,EAAE,CAAC,CAAC;wBAChB,MAAM,IAAI,CAAC,eAAe,CAAC,QAAQ,EAAE,IAAI,CAAC,GAAG,CAAC,SAAS,GAAG,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC;wBAClE,OAAO;oBACX,CAAC;oBAED,MAAM,IAAI,KAAK,CAAC,aAAa,QAAQ,cAAc,CAAC,CAAC;gBACzD,CAAC;YACL,CAAC;YACD,KAAK,CAAC,gBAAgB,CAAC,QAAiB,EAAE,UAAU,GAAG,KAAK;gBACxD,MAAM,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,eAAe,CAAC,IAAI,CAAC,CAAC;gBAE7C,IAAI,QAAQ,IAAI,CAAC,CAAC,QAAQ,CAAC,CAAC,GAAG,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;oBAC7C,MAAM,IAAI,KAAK,CAAC,aAAa,QAAQ,cAAc,CAAC,CAAC;gBACzD,CAAC;gBAED,OAAO,CAAC,CAAC;YACb,CAAC;SACJ,CAAC;IACN,CAAC;;AAqBL,gBAAgB;AAChB,SAAS,oBAAoB,CACzB,OAAuE;IAEvE,OAAO,CAAC,CAAE,OAA4C,CAAC,YAAY,CAAC;AACxE,CAAC;AAED,gBAAgB;AAChB,MAAM,CAAC,KAAK,UAAU,sBAAsB,CAAC,OAAuE;IAChH,MAAM,EAAE,OAAO,EAAE,mBAAmB,EAAE,MAAM,EAAE,kBAAkB,EAAE,eAAe,EAAE,GAAG,OAAO,CAAC;IAE9F,IAAI,CAAC,MAAM,EAAE,CAAC;QACV,MAAM,IAAI,KAAK,CAAC,0DAA0D,CAAC,CAAC;IAChF,CAAC;IAED,MAAM,OAAO,GAAG,sCAAsC,CAAC;QACnD,eAAe,EAAE,mBAAmB,EAAE,QAAQ;QAC9C,eAAe;QACf,kBAAkB;QAClB,mBAAmB,EAAE,mBAAmB,EAAE,OAAO;KACpD,CAAC,CAAC;IAEH,MAAM,IAAI,GAAG,qBAAqB,CAC9B,MAAM,EACN,mBAAmB,EAAE,QAAQ,IAAI,GAAG,EACpC,mBAAmB,EAAE,OAAO,IAAI,eAAe,IAAI,kBAAkB,CACxE,CAAC;IAEF,IAAI,oBAAoB,CAAC,OAAO,CAAC,EAAE,CAAC;QAChC,OAAO,OAAO,CAAC,YAAY,CAAC;YACxB,IAAI;YACJ,OAAO;YACP,GAAG,mBAAmB;SACzB,CAAC,CAAC;IACP,CAAC;IAED,OAAO,YAAY,CAAC;QAChB,YAAY,EAAE,OAAO,CAAC,YAAY;QAClC,aAAa,EAAE,OAAO,CAAC,aAAa;QACpC,gBAAgB,EAAE,OAAO,CAAC,gBAAgB;QAC1C,IAAI;QACJ,OAAO;QACP,GAAG,mBAAmB;KACzB,CAAC,CAAC;AACP,CAAC;AAED;;;GAGG;AACH,SAAS,qBAAqB,CAAC,MAAiB,EAAE,QAAgB,EAAE,OAAe;IAC/E,OAAO,KAAK,CAAC,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,gBAAgB,CAAC,QAAQ,CAAC,CAAC;SACxD,GAAG,CAAC,CAAC,CAAM,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;SACvB,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,KAAK,SAAS,IAAI,IAAI,KAAK,EAAE,CAAC;SACnD,GAAG,CAAC,CAAC,IAAwB,EAAE,EAAE;QAC9B,IAAI,IAAI,KAAK,SAAS,EAAE,CAAC;YACrB,OAAO,SAAS,CAAC;QACrB,CAAC;QACD,OAAO,cAAc,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC;IACzC,CAAC,CAAC;SACD,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,KAAK,SAAS,IAAI,IAAI,KAAK,EAAE,CAAa,CAAC;AACzE,CAAC;AAED;;;;;;;;;;;;;;;;;;;;;;;GAuBG;AACH,MAAM,UAAU,iBAAiB,CAG/B,MAAwC;IACtC,OAAO,MAAM,CAAC,MAAM,CAAU,MAAM,CAAC,CAAC;AAC1C,CAAC"}