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

package/internals/adaptive-playwright-crawler.d.ts

@@ -3,8 +3,7 @@ import type { BasicCrawlerOptions } from '@crawlee/basic';
 import { BasicCrawler } from '@crawlee/basic';
 import type { ContextPipeline, CrawlingContext, EnqueueLinksOptions, GetUserDataFromRequest, RouterRoutes, StorageTransactionView } from '@crawlee/core';
 import type { Dictionary, Awaitable } from '@crawlee/types';
-import { type CheerioRoot } from '@crawlee/utils/internal';
-import { type Cheerio } from 'cheerio';
+import { type Cheerio, type CheerioAPI } from 'cheerio';
 import type { AnyNode } from 'domhandler';
 // @ts-ignore optional peer dependency or compatibility with es2022
 import type { Page } from 'playwright';
@@ -80,7 +79,7 @@ export interface AdaptivePlaywrightCrawlerContext<UserData extends Dictionary =
      * });
      * ```
      */
-    parseWithCheerio(selector?: string, timeoutMs?: number): Promise<CheerioRoot>;
+    parseWithCheerio(selector?: string, timeoutMs?: number): Promise<CheerioAPI>;
     enqueueLinks(options?: EnqueueLinksOptions): Promise<unknown>;
 }
 interface AdaptiveHookContext extends Pick<AdaptivePlaywrightCrawlerContext, 'id' | 'session' | 'proxyInfo' | 'log'> {
@@ -186,16 +185,7 @@ export declare class AdaptivePlaywrightCrawler<ContextExtension = Dictionary<nev
     #private;
     constructor(options?: AdaptivePlaywrightCrawlerOptions<ContextExtension, ExtendedContext, Routes, StatisticStateExtension>);
     protected init(): Promise<void>;
-    protected buildContextPipeline(): ContextPipeline<CrawlingContext<Dictionary>, CrawlingContext<Dictionary> & {
-        readonly request: LoadedRequest<Request<Dictionary>>;
-        readonly response: Response;
-        readonly page: Page;
-        readonly querySelector: AdaptivePlaywrightCrawlerContext["querySelector"];
-        readonly querySelectorAll: AdaptivePlaywrightCrawlerContext["querySelectorAll"];
-        readonly waitForSelector: AdaptivePlaywrightCrawlerContext["waitForSelector"];
-        readonly parseWithCheerio: AdaptivePlaywrightCrawlerContext["parseWithCheerio"];
-        readonly enqueueLinks: AdaptivePlaywrightCrawlerContext["enqueueLinks"];
-    }>;
+    protected buildContextPipeline(): ContextPipeline<CrawlingContext, AdaptivePlaywrightCrawlerContext>;
     private adaptCheerioContext;
     private adaptPlaywrightContext;
     /**

package/internals/adaptive-playwright-crawler.js

@@ -2,8 +2,8 @@ import { isDeepStrictEqual } from 'node:util';
 import { BasicCrawler } from '@crawlee/basic';
 import { extractUrlsFromPage } from '@crawlee/browser';
 import { CheerioCrawler } from '@crawlee/cheerio';
-import { createStorageTransaction, EnqueueStrategy, OwnedOrInjected, parseArgument, RequestHandlerError, resolveBaseUrlForEnqueueLinksFiltering, Router, Statistics, } from '@crawlee/core';
-import { extractUrlsFromCheerio } from '@crawlee/utils/internal';
+import { createStorageTransaction, EnqueueStrategy, OwnedOrInjected, RequestHandlerError, resolveBaseUrlForEnqueueLinksFiltering, Router, Statistics, } from '@crawlee/core';
+import { extractUrlsFromCheerio, parseArgument } from '@crawlee/utils/internal';
 import { z } from 'zod';
 import { addTimeoutToPromise } from '@apify/timeout';
 import { PlaywrightCrawler } from './playwright-crawler.js';

package/internals/enqueue-links/click-elements.js

@@ -1,5 +1,6 @@
 import { URL } from 'node:url';
-import { applyRequestTransform, constructUrlPatternObjects, createRequestOptions, filterRequestOptionsByPatterns, parseArgument, urlPatternSchema, Request as CrawleeRequest, schemas, serviceLocator, } from '@crawlee/browser';
+import { applyRequestTransform, constructUrlPatternObjects, createRequestOptions, filterRequestOptionsByPatterns, urlPatternSchema, Request as CrawleeRequest, serviceLocator, } from '@crawlee/browser';
+import { parseArgument, schemas } from '@crawlee/utils/internal';
 import { z } from 'zod';
 const STARTING_Z_INDEX = 2147400000;
 const getLog = () => serviceLocator.getChildLog('Playwright Click Elements');

package/internals/playwright-crawler.d.ts

@@ -1,12 +1,11 @@
-import type { BrowserCrawlerOptions, BrowserCrawlingContext, BrowserHook, GetUserDataFromRequest, RequestHandler, RouterHandler, RouterRoutes, RouteSchemas, RoutesFromSchemas } from '@crawlee/browser';
+import type { BrowserCrawlerOptions, BrowserCrawlingContext, BrowserHook, ContextPipeline, CrawlingContext, GetUserDataFromRequest, RequestHandler, RouterHandler, RouterRoutes, RouteSchemas, RoutesFromSchemas } from '@crawlee/browser';
 import { BrowserCrawler } from '@crawlee/browser';
 import type { Dictionary } from '@crawlee/types';
 // @ts-ignore optional peer dependency or compatibility with es2022
-import type { Download, LaunchOptions, Page, Response } from 'playwright';
+import type { LaunchOptions, Page, Response } from 'playwright';
 import { z } from 'zod';
-import type { EnqueueLinksByClickingElementsOptions } from './enqueue-links/click-elements.js';
 import type { PlaywrightLaunchContext } from './playwright-launcher.js';
-import type { BlockRequestsOptions, DirectNavigationOptions, HandleCloudflareChallengeOptions, InfiniteScrollOptions, InjectFileOptions, PlaywrightContextUtils, SaveSnapshotOptions } from './utils/playwright-utils.js';
+import type { DirectNavigationOptions, HandleCloudflareChallengeOptions, PlaywrightContextUtils } from './utils/playwright-utils.js';
 export type PlaywrightGotoOptions = NonNullable<Parameters<Page['goto']>[1]>;
 export interface PlaywrightCrawlingContext<UserData extends Dictionary = any> extends BrowserCrawlingContext<Page, Response, UserData, PlaywrightGotoOptions>, PlaywrightContextUtils {
 }
@@ -145,6 +144,9 @@ export interface PlaywrightCrawlerOptions<ContextExtension = Dictionary<never>,
  * @category Crawlers
  */
 export declare class PlaywrightCrawler<ContextExtension = Dictionary<never>, ExtendedContext extends PlaywrightCrawlingContext = PlaywrightCrawlingContext & ContextExtension, Routes extends Record<keyof Routes, Dictionary> = Record<string, GetUserDataFromRequest<PlaywrightCrawlingContext['request']>>, StatisticStateExtension extends object = {}> extends BrowserCrawler<Page, Response, LaunchOptions, PlaywrightCrawlingContext, ContextExtension, ExtendedContext, Routes, StatisticStateExtension> {
+    /**
+     * @internal
+     */
     protected static optionsShape: {
         headless: z.ZodOptional<z.ZodBoolean>;
         launcher: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
@@ -204,6 +206,7 @@ export declare class PlaywrightCrawler<ContextExtension = Dictionary<never>, Ext
         statistics: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
         id: z.ZodOptional<z.ZodString>;
     };
+    /** @internal */
     protected static optionsSchema: z.ZodObject<{
         headless: z.ZodOptional<z.ZodBoolean>;
         launcher: z.ZodOptional<z.ZodCustom<Dictionary, Dictionary>>;
@@ -267,24 +270,7 @@ export declare class PlaywrightCrawler<ContextExtension = Dictionary<never>, Ext
      * All `PlaywrightCrawler` parameters are passed via an options object.
      */
     constructor(options?: PlaywrightCrawlerOptions<ContextExtension, ExtendedContext, Routes, StatisticStateExtension>);
-// @ts-ignore optional peer dependency or compatibility with es2022
-    protected buildContextPipeline(): import("@crawlee/browser").ContextPipeline<import("@crawlee/browser").CrawlingContext<Dictionary>, BrowserCrawlingContext<Page, Response, Dictionary, Dictionary> & {
-        injectFile: (filePath: string, options?: InjectFileOptions) => Promise<unknown>;
-        injectJQuery: () => Promise<void>;
-        blockRequests: (options?: BlockRequestsOptions) => Promise<void>;
-        waitForSelector: (selector: string, timeoutMs?: number) => Promise<void>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        parseWithCheerio: (selector?: string, timeoutMs?: number) => Promise<import("@crawlee/browser").CheerioAPI>;
-        infiniteScroll: (options?: InfiniteScrollOptions) => Promise<void>;
-        listDownloads: () => Promise<Download[]>;
-        saveSnapshot: (options?: SaveSnapshotOptions) => Promise<void>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        enqueueLinksByClickingElements: (options: Omit<EnqueueLinksByClickingElementsOptions, "page" | "requestManager">) => Promise<import("@crawlee/types").BatchAddRequestsResult>;
-// @ts-ignore optional peer dependency or compatibility with es2022
-        compileScript: (scriptString: string, ctx?: Dictionary) => import("./utils/playwright-utils.js").CompiledScriptFunction;
-        closeCookieModals: () => Promise<void>;
-        handleCloudflareChallenge: (options?: HandleCloudflareChallengeOptions) => Promise<Response | undefined>;
-    }>;
+    protected buildContextPipeline(): ContextPipeline<CrawlingContext, PlaywrightCrawlingContext>;
     protected navigationHandler(crawlingContext: PlaywrightCrawlingContext, gotoOptions: DirectNavigationOptions): Promise<Response | null>;
     private enhanceContext;
 }

package/internals/playwright-crawler.js

@@ -1,4 +1,5 @@
-import { assertBrowserPoolNotConfigured, BrowserCrawler, parseArgument, RequestState, Router, schemas, serviceLocator, } from '@crawlee/browser';
+import { assertBrowserPoolNotConfigured, BrowserCrawler, RequestState, Router, serviceLocator } from '@crawlee/browser';
+import { parseArgument, schemas } from '@crawlee/utils/internal';
 import { z } from 'zod';
 import { playwrightBrowserPool, remotePlaywrightBrowserPool } from './playwright-browser-pool.js';
 import { gotoExtended, playwrightUtils } from './utils/playwright-utils.js';
@@ -68,11 +69,15 @@ import { gotoExtended, playwrightUtils } from './utils/playwright-utils.js';
  * @category Crawlers
  */
 export class PlaywrightCrawler extends BrowserCrawler {
+    /**
+     * @internal
+     */
     static optionsShape = {
         ...BrowserCrawler.optionsShape,
         headless: z.boolean().optional(),
         launcher: schemas.anyObject.optional(),
     };
+    /** @internal */
     static optionsSchema = z.strictObject(PlaywrightCrawler.optionsShape);
     /**
      * All `PlaywrightCrawler` parameters are passed via an options object.

package/internals/playwright-launcher.d.ts

@@ -72,6 +72,9 @@ export interface PlaywrightLaunchContext extends BrowserLaunchContext<LaunchOpti
  */
 export declare class PlaywrightLauncher extends BrowserLauncher<PlaywrightPlugin> {
     readonly configuration: Configuration;
+    /**
+     * @internal
+     */
     protected static optionsShape: {
 // @ts-ignore optional peer dependency or compatibility with es2022
         launcher: z.ZodOptional<z.ZodCustom<import("@crawlee/types").Dictionary, import("@crawlee/types").Dictionary>>;
@@ -87,6 +90,7 @@ export declare class PlaywrightLauncher extends BrowserLauncher<PlaywrightPlugin
         launchOptions: z.ZodOptional<z.ZodCustom<import("@crawlee/types").Dictionary, import("@crawlee/types").Dictionary>>;
         userAgent: z.ZodOptional<z.ZodString>;
     };
+    /** @internal */
     protected static optionsSchema: z.ZodObject<{
 // @ts-ignore optional peer dependency or compatibility with es2022
         launcher: z.ZodOptional<z.ZodCustom<import("@crawlee/types").Dictionary, import("@crawlee/types").Dictionary>>;

package/internals/playwright-launcher.js

@@ -1,5 +1,6 @@
-import { BrowserLauncher, Configuration, parseArgument, schemas } from '@crawlee/browser';
+import { BrowserLauncher, Configuration } from '@crawlee/browser';
 import { PlaywrightPlugin } from '@crawlee/browser-pool';
+import { parseArgument, schemas } from '@crawlee/utils/internal';
 import { z } from 'zod';
 /**
  * `PlaywrightLauncher` is based on the `BrowserLauncher`. It launches `playwright` browser instance.
@@ -7,12 +8,16 @@ import { z } from 'zod';
  */
 export class PlaywrightLauncher extends BrowserLauncher {
     configuration;
+    /**
+     * @internal
+     */
     static optionsShape = {
         ...BrowserLauncher.optionsShape,
         // Passthrough schemas — the launcher module object must keep its prototype through parsing.
         launcher: schemas.anyObject.optional(),
         launchContextOptions: schemas.anyObject.optional(),
     };
+    /** @internal */
     static optionsSchema = z.strictObject(PlaywrightLauncher.optionsShape);
     /**
      * All `PlaywrightLauncher` parameters are passed via this launchContext object.

package/internals/utils/playwright-utils.d.ts

@@ -19,7 +19,7 @@
  */
 import { Configuration, type Request } from '@crawlee/browser';
 import type { BatchAddRequestsResult, Dictionary } from '@crawlee/types';
-import { type CheerioRoot } from '@crawlee/utils/internal';
+import type { CheerioAPI } from 'cheerio';
 // @ts-ignore optional peer dependency or compatibility with es2022
 import type { Download, Page, Response } from 'playwright';
 import type { EnqueueLinksByClickingElementsOptions } from '../enqueue-links/click-elements.js';
@@ -287,7 +287,7 @@ export declare function saveSnapshot(page: Page, options?: SaveSnapshotOptions):
  * @param page Playwright [`Page`](https://playwright.dev/docs/api/class-page) object.
  * @param ignoreShadowRoots
  */
-export declare function parseWithCheerio(page: Page, ignoreShadowRoots?: boolean, ignoreIframes?: boolean): Promise<CheerioRoot>;
+export declare function parseWithCheerio(page: Page, ignoreShadowRoots?: boolean, ignoreIframes?: boolean): Promise<CheerioAPI>;
 export declare function closeCookieModals(page: Page): Promise<void>;
 export interface HandleCloudflareChallengeOptions {
     /** Logging defaults to the `debug` level, use this flag to log to `info` level instead. */
@@ -434,7 +434,7 @@ export interface PlaywrightContextUtils {
      * });
      * ```
      */
-    parseWithCheerio(selector?: string, timeoutMs?: number): Promise<CheerioRoot>;
+    parseWithCheerio(selector?: string, timeoutMs?: number): Promise<CheerioAPI>;
     /**
      * Scrolls to the bottom of a page, or until it times out.
      * Loads dynamic content when it hits the bottom of a page, and then continues scrolling.

package/internals/utils/playwright-utils.js

@@ -20,8 +20,9 @@
 import { readFile } from 'node:fs/promises';
 import { createRequire } from 'node:module';
 import vm from 'node:vm';
-import { Configuration, KeyValueStore, parseArgument, schemas, serviceLocator, SessionError, validators, } from '@crawlee/browser';
+import { Configuration, KeyValueStore, serviceLocator, SessionError, validators } from '@crawlee/browser';
 import { expandShadowRoots, sleep } from '@crawlee/utils';
+import { parseArgument, schemas } from '@crawlee/utils/internal';
 import { z } from 'zod';
 import { LruCache } from '@apify/datastructures';
 import { enqueueLinksByClickingElements } from '../enqueue-links/click-elements.js';

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@crawlee/playwright",
-    "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"
@@ -49,13 +49,13 @@
     "dependencies": {
         "@apify/datastructures": "^2.0.3",
         "@apify/timeout": "^0.4.4",
-        "@crawlee/basic": "4.0.0-beta.134",
-        "@crawlee/browser": "4.0.0-beta.134",
-        "@crawlee/browser-pool": "4.0.0-beta.134",
-        "@crawlee/cheerio": "4.0.0-beta.134",
-        "@crawlee/core": "4.0.0-beta.134",
-        "@crawlee/types": "4.0.0-beta.134",
-        "@crawlee/utils": "4.0.0-beta.134",
+        "@crawlee/basic": "4.0.0-beta.135",
+        "@crawlee/browser": "4.0.0-beta.135",
+        "@crawlee/browser-pool": "4.0.0-beta.135",
+        "@crawlee/cheerio": "4.0.0-beta.135",
+        "@crawlee/core": "4.0.0-beta.135",
+        "@crawlee/types": "4.0.0-beta.135",
+        "@crawlee/utils": "4.0.0-beta.135",
         "cheerio": "^1.0.0",
         "jquery": "^3.7.1",
         "ml-logistic-regression": "^2.0.0",
@@ -84,5 +84,5 @@
             }
         }
     },
-    "gitHead": "0f269d5b24ec2829f3d44c19a35973fa20d95999"
+    "gitHead": "ca0325880bbcc753d97506796e9bda3b35bca201"
 }