@crawlee/jsdom 4.0.0-beta.134 → 4.0.0-beta.135

package/internals/jsdom-crawler.d.ts

@@ -1,7 +1,7 @@
-import type { AddRequestsBatchedResult, CrawlingContext, EnqueueLinksOptions, ErrorHandler, ExtractLinksOptions, GetUserDataFromRequest, HttpCrawlerOptions, InternalHttpCrawlingContext, InternalHttpHook, RequestHandler, RouterHandler, RouterRoutes, RouteSchemas, RoutesFromSchemas } 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 } from '@crawlee/utils/internal';
+import type { CheerioAPI } from 'cheerio';
 import type { DOMWindow } from 'jsdom';
 import { VirtualConsole } from 'jsdom';
 import { z } from 'zod';
@@ -53,7 +53,7 @@ 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.
      */
@@ -67,6 +67,9 @@ export type JSDOMRequestHandler<UserData extends Dictionary = any, // with defau
 JSONData extends Dictionary = any> = RequestHandler<JSDOMCrawlingContext<UserData, JSONData>>;
 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: {
         runScripts: z.ZodOptional<z.ZodBoolean>;
         hideInternalConsole: z.ZodOptional<z.ZodBoolean>;
@@ -124,6 +127,7 @@ export declare class JSDOMCrawler<ContextExtension = Dictionary<never>, Extended
         statistics: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
         id: z.ZodOptional<z.ZodString>;
     };
+    /** @internal */
     protected static optionsSchema: z.ZodObject<{
         runScripts: z.ZodOptional<z.ZodBoolean>;
         hideInternalConsole: z.ZodOptional<z.ZodBoolean>;
@@ -182,18 +186,7 @@ export declare class JSDOMCrawler<ContextExtension = Dictionary<never>, Extended
         id: z.ZodOptional<z.ZodString>;
     }, z.core.$strict>;
     constructor(options?: JSDOMCrawlerOptions<ContextExtension, ExtendedContext, any, any, Routes, StatisticStateExtension>);
-// @ts-ignore optional peer dependency or compatibility with es2022
-    protected buildContextPipeline(): import("@crawlee/http").ContextPipeline<CrawlingContext<Dictionary>, InternalHttpCrawlingContext<any, any> & {
-        readonly window: DOMWindow;
-        readonly body: string;
-        readonly document: Document;
-    } & {
-        extractLinks: (options?: ExtractLinksOptions) => Promise<string[]>;
-        enqueueLinks: (options?: EnqueueLinksOptions) => Promise<AddRequestsBatchedResult>;
-        waitForSelector(selector: string, timeoutMs?: number): Promise<void>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        parseWithCheerio(selector?: string, _timeoutMs?: number): Promise<import("cheerio").CheerioAPI>;
-    }>;
+    protected buildContextPipeline(): ContextPipeline<CrawlingContext, JSDOMCrawlingContext>;
     /**
      * Returns the currently used `VirtualConsole` instance. Can be used to listen for the JSDOM's internal console messages.
      *

package/internals/jsdom-crawler.js

@@ -1,5 +1,6 @@
-import { EnqueueStrategy, HttpCrawler, NavigationSkippedError, parseArgument, resolveBaseUrlForEnqueueLinksFiltering, Router, tryAbsoluteURL, } from '@crawlee/http';
+import { EnqueueStrategy, HttpCrawler, NavigationSkippedError, resolveBaseUrlForEnqueueLinksFiltering, Router, } from '@crawlee/http';
 import { sleep } from '@crawlee/utils';
+import { parseArgument, tryAbsoluteURL } from '@crawlee/utils/internal';
 import { JSDOM, ResourceLoader, VirtualConsole } from 'jsdom';
 import { z } from 'zod';
 import { addTimeoutToPromise } from '@apify/timeout';
@@ -84,11 +85,15 @@ 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: z.boolean().optional(),
         hideInternalConsole: z.boolean().optional(),
     };
+    /** @internal */
     static optionsSchema = z.strictObject(JSDOMCrawler.optionsShape);
     #runScripts;
     #hideInternalConsole;

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@crawlee/jsdom",
-    "version": "4.0.0-beta.134",
+    "version": "4.0.0-beta.135",
     "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"
@@ -49,9 +49,9 @@
     "dependencies": {
         "@apify/timeout": "^0.4.4",
         "@apify/utilities": "^2.7.10",
-        "@crawlee/http": "4.0.0-beta.134",
-        "@crawlee/types": "4.0.0-beta.134",
-        "@crawlee/utils": "4.0.0-beta.134",
+        "@crawlee/http": "4.0.0-beta.135",
+        "@crawlee/types": "4.0.0-beta.135",
+        "@crawlee/utils": "4.0.0-beta.135",
         "@types/jsdom": "^21.1.7",
         "cheerio": "^1.0.0",
         "jsdom": "^26.1.0",
@@ -65,5 +65,5 @@
             }
         }
     },
-    "gitHead": "0f269d5b24ec2829f3d44c19a35973fa20d95999"
+    "gitHead": "ca0325880bbcc753d97506796e9bda3b35bca201"
 }