apify--crawlee/docs/public-api/crawlee-cheerio.api.md

API Report File for "@crawlee/cheerio"

Do not edit this file. It is a report generated by API Extractor.

import type { BasicCrawlingContext } from '@crawlee/http';
import type { BatchAddRequestsResult } from '@crawlee/types';
import * as cheerio from 'cheerio';
import type { CheerioAPI } from 'cheerio';
import { CheerioRoot } from '@crawlee/utils';
import { ContextPipeline } from '@crawlee/http';
import { CrawlingContext } from '@crawlee/http';
import type { Dictionary } from '@crawlee/types';
import type { EnqueueLinksOptions } from '@crawlee/http';
import type { ErrorHandler } from '@crawlee/http';
import type { GetUserDataFromRequest } from '@crawlee/http';
import { HttpCrawler } from '@crawlee/http';
import type { HttpCrawlerOptions } from '@crawlee/http';
import type { InternalHttpCrawlingContext } from '@crawlee/http';
import type { InternalHttpHook } from '@crawlee/http';
import { IRequestManager } from '@crawlee/http';
import type { RequestHandler } from '@crawlee/http';
import { RobotsTxtFile } from '@crawlee/utils';
import { RouterHandler } from '@crawlee/http';
import type { RouterRoutes } from '@crawlee/http';
import type { SkippedRequestCallback } from '@crawlee/http';

// @public
export class CheerioCrawler<ContextExtension = Dictionary<never>, ExtendedContext extends CheerioCrawlingContext = CheerioCrawlingContext & ContextExtension> extends HttpCrawler<CheerioCrawlingContext, ContextExtension, ExtendedContext> {
    constructor(options?: CheerioCrawlerOptions<ContextExtension, ExtendedContext>);
    // (undocumented)
    protected buildContextPipeline(): ContextPipeline<CrawlingContext<Dictionary>, InternalHttpCrawlingContext<any, any> & {
    readonly body: string;
    readonly $: CheerioAPI;
    } & {
    enqueueLinks: (enqueueOptions?: EnqueueLinksOptions) => Promise<BatchAddRequestsResult>;
    waitForSelector: (selector: string, _timeoutMs?: number) => Promise<void>;
    parseWithCheerio: (selector?: string, timeoutMs?: number) => Promise<CheerioAPI>;
    }>;
}

// @internal (undocumented)
export function cheerioCrawlerEnqueueLinks(options: EnqueueLinksInternalOptions | BoundEnqueueLinksInternalOptions): Promise<unknown>;

// @public (undocumented)
export interface CheerioCrawlerOptions<ContextExtension = Dictionary<never>, ExtendedContext extends CheerioCrawlingContext = CheerioCrawlingContext & ContextExtension, UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
JSONData extends Dictionary = any> extends HttpCrawlerOptions<CheerioCrawlingContext<UserData, JSONData>, ContextExtension, ExtendedContext> {
}

// @public (undocumented)
export interface CheerioCrawlingContext<UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
JSONData extends Dictionary = any> extends InternalHttpCrawlingContext<UserData, JSONData> {
    $: cheerio.CheerioAPI;
    body: string;
    enqueueLinks(options?: EnqueueLinksOptions): Promise<BatchAddRequestsResult>;
    parseWithCheerio(selector?: string, timeoutMs?: number): Promise<CheerioRoot>;
    waitForSelector(selector: string, timeoutMs?: number): Promise<void>;
}

// @public (undocumented)
export type CheerioErrorHandler<UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
JSONData extends Dictionary = any> = ErrorHandler<CheerioCrawlingContext<UserData, JSONData>>;

// @public (undocumented)
export type CheerioHook<UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
JSONData extends Dictionary = any> = InternalHttpHook<CheerioCrawlingContext<UserData, JSONData>>;

// @public (undocumented)
export type CheerioRequestHandler<UserData extends Dictionary = any, // with default to Dictionary we cant use a typed router in untyped crawler
JSONData extends Dictionary = any> = RequestHandler<CheerioCrawlingContext<UserData, JSONData>>;

// @public
export function createCheerioRouter<Context extends CheerioCrawlingContext = CheerioCrawlingContext, UserData extends Dictionary = GetUserDataFromRequest<Context['request']>>(routes?: RouterRoutes<Context, UserData>): RouterHandler<Context>;


export * from "@crawlee/http";

// (No @packageDocumentation comment for this package)