@crawlee/basic 4.0.0-beta.2 → 4.0.0-beta.21

package/README.md

@@ -9,6 +9,10 @@
     <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>
+
 <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>
@@ -24,7 +28,7 @@ Crawlee is available as the [`crawlee`](https://www.npmjs.com/package/crawlee) N
 
 > 👉 **View full documentation, guides and examples on the [Crawlee project website](https://crawlee.dev)** 👈
 
-> Crawlee for Python is open for early adopters. 🐍  [👉 Checkout the source code 👈](https://github.com/apify/crawlee-python).
+> Do you prefer 🐍 Python instead of JavaScript? [👉 Checkout Crawlee for Python 👈](https://github.com/apify/crawlee-python).
 
 ## Installation
 

package/index.d.ts

@@ -1,4 +1,4 @@
 export * from '@crawlee/core';
 export * from './internals/basic-crawler.js';
-export * from './internals/constants.js';
+export { CheerioRoot, CheerioAPI, Cheerio, Element } from '@crawlee/utils';
 //# sourceMappingURL=index.d.ts.map

package/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,cAAc,eAAe,CAAC;AAC9B,cAAc,8BAA8B,CAAC;AAC7C,cAAc,0BAA0B,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,cAAc,eAAe,CAAC;AAC9B,cAAc,8BAA8B,CAAC;AAC7C,OAAO,EAAE,WAAW,EAAE,UAAU,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,gBAAgB,CAAC"}

package/index.js

@@ -1,4 +1,3 @@
 export * from '@crawlee/core';
 export * from './internals/basic-crawler.js';
-export * from './internals/constants.js';
 //# sourceMappingURL=index.js.map

package/index.js.map

@@ -1 +1 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,cAAc,eAAe,CAAC;AAC9B,cAAc,8BAA8B,CAAC;AAC7C,cAAc,0BAA0B,CAAC"}
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,cAAc,eAAe,CAAC;AAC9B,cAAc,8BAA8B,CAAC"}

package/internals/basic-crawler.d.ts

@@ -1,38 +1,14 @@
-import type { AddRequestsBatchedOptions, AddRequestsBatchedResult, AutoscaledPoolOptions, BaseHttpClient, CrawlingContext, DatasetExportOptions, EnqueueLinksOptions, EventManager, FinalStatistics, GetUserDataFromRequest, IRequestList, LoadedContext, ProxyInfo, Request, RequestOptions, RestrictedCrawlingContext, RouterHandler, RouterRoutes, Session, SessionPoolOptions, SkippedRequestCallback, Source, StatisticsOptions, StatisticState } from '@crawlee/core';
-import { AutoscaledPool, Configuration, Dataset, RequestProvider, SessionPool, Statistics } from '@crawlee/core';
-import type { Awaitable, BatchAddRequestsResult, Dictionary, SetStatusMessageOptions } from '@crawlee/types';
+import type { AddRequestsBatchedOptions, AddRequestsBatchedResult, AutoscaledPoolOptions, CrawlingContext, DatasetExportOptions, EnqueueLinksOptions, EventManager, FinalStatistics, GetUserDataFromRequest, IRequestList, IRequestManager, ProxyConfiguration, Request, RequestsLike, RouterHandler, RouterRoutes, Session, SessionPoolOptions, SkippedRequestCallback, Source, StatisticsOptions, StatisticState } from '@crawlee/core';
+import { AutoscaledPool, Configuration, ContextPipeline, Dataset, RequestProvider, SessionPool, Statistics } from '@crawlee/core';
+import type { Awaitable, BaseHttpClient, BatchAddRequestsResult, Dictionary, ProxyInfo, SetStatusMessageOptions } from '@crawlee/types';
 import { RobotsTxtFile } from '@crawlee/utils';
-import type { SetRequired } from 'type-fest';
+import type { ReadonlyDeep, SetRequired } from 'type-fest';
 import type { Log } from '@apify/log';
 import { TimeoutError } from '@apify/timeout';
-export interface BasicCrawlingContext<UserData extends Dictionary = Dictionary> extends CrawlingContext<BasicCrawler, UserData> {
-    /**
-     * This function automatically finds and enqueues links from the current page, adding them to the {@link RequestQueue}
-     * currently used by the crawler.
-     *
-     * Optionally, the function allows you to filter the target links' URLs using an array of globs or regular expressions
-     * and override settings of the enqueued {@link Request} objects.
-     *
-     * Check out the [Crawl a website with relative links](https://crawlee.dev/js/docs/examples/crawl-relative-links) example
-     * for more details regarding its usage.
-     *
-     * **Example usage**
-     *
-     * ```ts
-     * async requestHandler({ enqueueLinks }) {
-     *     await enqueueLinks({
-     *       urls: [...],
-     *     });
-     * },
-     * ```
-     *
-     * @param [options] All `enqueueLinks()` parameters are passed via an options object.
-     * @returns Promise that resolves to {@link BatchAddRequestsResult} object.
-     */
-    enqueueLinks(options?: SetRequired<EnqueueLinksOptions, 'urls'>): Promise<BatchAddRequestsResult>;
+export interface BasicCrawlingContext<UserData extends Dictionary = Dictionary> extends CrawlingContext<UserData> {
 }
-export type RequestHandler<Context extends CrawlingContext = LoadedContext<BasicCrawlingContext & RestrictedCrawlingContext>> = (inputs: LoadedContext<Context>) => Awaitable<void>;
-export type ErrorHandler<Context extends CrawlingContext = LoadedContext<BasicCrawlingContext & RestrictedCrawlingContext>> = (inputs: LoadedContext<Context>, error: Error) => Awaitable<void>;
+export type RequestHandler<Context extends CrawlingContext = CrawlingContext> = (inputs: Context) => Awaitable<void>;
+export type ErrorHandler<Context extends CrawlingContext = CrawlingContext, ExtendedContext extends Context = Context> = (inputs: Context & Partial<ExtendedContext>, error: Error) => Awaitable<void>;
 export interface StatusMessageCallbackParams<Context extends CrawlingContext = BasicCrawlingContext, Crawler extends BasicCrawler<any> = BasicCrawler<Context>> {
     state: StatisticState;
     crawler: Crawler;
@@ -40,7 +16,10 @@ export interface StatusMessageCallbackParams<Context extends CrawlingContext = B
     message: string;
 }
 export type StatusMessageCallback<Context extends CrawlingContext = BasicCrawlingContext, Crawler extends BasicCrawler<any> = BasicCrawler<Context>> = (params: StatusMessageCallbackParams<Context, Crawler>) => Awaitable<void>;
-export interface BasicCrawlerOptions<Context extends CrawlingContext = BasicCrawlingContext> {
+export type RequireContextPipeline<DefaultContextType extends CrawlingContext, FinalContextType extends DefaultContextType> = DefaultContextType extends FinalContextType ? {} : {
+    contextPipelineBuilder: () => ContextPipeline<CrawlingContext, FinalContextType>;
+};
+export interface BasicCrawlerOptions<Context extends CrawlingContext = CrawlingContext, ContextExtension = Dictionary<never>, ExtendedContext extends Context = Context & ContextExtension> {
     /**
      * User-provided function that performs the logic of the crawler. It is called for each URL to crawl.
      *
@@ -58,7 +37,35 @@ export interface BasicCrawlerOptions<Context extends CrawlingContext = BasicCraw
      * The exceptions are logged to the request using the
      * {@link Request.pushErrorMessage|`Request.pushErrorMessage()`} function.
      */
-    requestHandler?: RequestHandler<LoadedContext<Context>>;
+    requestHandler?: RequestHandler<ExtendedContext>;
+    /**
+     * Allows the user to extend the crawling context passed to the request handler with custom functionality.
+     *
+     * **Example usage:**
+     *
+     * ```javascript
+     * import { BasicCrawler } from 'crawlee';
+     *
+     * // Create a crawler instance
+     * const crawler = new BasicCrawler({
+     *     extendContext(context) => ({
+     *         async customHelper() {
+     *             await context.pushData({ url: context.request.url })
+     *         }
+     *     }),
+     *     async requestHandler(context) {
+     *         await context.customHelper();
+     *     },
+     * });
+     * ```
+     */
+    extendContext?: (context: Context) => Awaitable<ContextExtension>;
+    /**
+     * *Intended for BasicCrawler subclasses*. Prepares a context pipeline that transforms the initial crawling context into the shape given by the `Context` type parameter.
+     *
+     * The option is not required if your crawler subclass does not extend the crawling context with custom information or helpers.
+     */
+    contextPipelineBuilder?: () => ContextPipeline<CrawlingContext, Context>;
     /**
      * Static list of URLs to be processed.
      * If not provided, the crawler will open the default request queue when the {@link BasicCrawler.addRequests|`crawler.addRequests()`} function is called.
@@ -73,6 +80,13 @@ export interface BasicCrawlerOptions<Context extends CrawlingContext = BasicCraw
      * it is a shortcut for running `crawler.addRequests()` before the `crawler.run()`.
      */
     requestQueue?: RequestProvider;
+    /**
+     * Allows explicitly configuring a request manager. Mutually exclusive with the `requestQueue` and `requestList` options.
+     *
+     * This enables explicitly configuring the crawler to use `RequestManagerTandem`, for instance.
+     * If using this, the type of `BasicCrawler.requestQueue` may not be fully compatible with the `RequestProvider` class.
+     */
+    requestManager?: IRequestManager;
     /**
      * Timeout in which the function passed as {@link BasicCrawlerOptions.requestHandler|`requestHandler`} needs to finish, in seconds.
      * @default 60
@@ -87,7 +101,7 @@ export interface BasicCrawlerOptions<Context extends CrawlingContext = BasicCraw
      * Second argument is the `Error` instance that
      * represents the last error thrown during processing of the request.
      */
-    errorHandler?: ErrorHandler<Context>;
+    errorHandler?: ErrorHandler<CrawlingContext, ExtendedContext>;
     /**
      * A function to handle requests that failed more than {@link BasicCrawlerOptions.maxRequestRetries|`maxRequestRetries`} times.
      *
@@ -96,7 +110,7 @@ export interface BasicCrawlerOptions<Context extends CrawlingContext = BasicCraw
      * Second argument is the `Error` instance that
      * represents the last error thrown during processing of the request.
      */
-    failedRequestHandler?: ErrorHandler<Context>;
+    failedRequestHandler?: ErrorHandler<CrawlingContext, ExtendedContext>;
     /**
      * Specifies the maximum number of retries allowed for a request if its processing fails.
      * This includes retries due to navigation errors or errors thrown from user-supplied functions
@@ -126,12 +140,18 @@ export interface BasicCrawlerOptions<Context extends CrawlingContext = BasicCraw
      * > *NOTE:* In cases of parallel crawling, the actual number of pages visited might be slightly higher than this value.
      */
     maxRequestsPerCrawl?: number;
+    /**
+     * Maximum depth of the crawl. If not set, the crawl will continue until all requests are processed.
+     * Setting this to `0` will only process the initial requests, skipping all links enqueued by `crawlingContext.enqueueLinks` and `crawlingContext.addRequests`.
+     * Passing `1` will process the initial requests and all links enqueued by `crawlingContext.enqueueLinks` and `crawlingContext.addRequests` in the handler for initial requests.
+     */
+    maxCrawlDepth?: number;
     /**
      * Custom options passed to the underlying {@link AutoscaledPool} constructor.
      * > *NOTE:* The {@link AutoscaledPoolOptions.runTaskFunction|`runTaskFunction`}
-     * and {@link AutoscaledPoolOptions.isTaskReadyFunction|`isTaskReadyFunction`} options
-     * are provided by the crawler and cannot be overridden.
-     * However, we can provide a custom implementation of {@link AutoscaledPoolOptions.isFinishedFunction|`isFinishedFunction`}.
+     * option is provided by the crawler and cannot be overridden.
+     * However, we can provide custom implementations of {@link AutoscaledPoolOptions.isFinishedFunction|`isFinishedFunction`}
+     * and {@link AutoscaledPoolOptions.isTaskReadyFunction|`isTaskReadyFunction`}.
      */
     autoscaledPoolOptions?: AutoscaledPoolOptions;
     /**
@@ -199,11 +219,20 @@ export interface BasicCrawlerOptions<Context extends CrawlingContext = BasicCraw
     /**
      * If set to `true`, the crawler will automatically try to fetch the robots.txt file for each domain,
      * and skip those that are not allowed. This also prevents disallowed URLs to be added via `enqueueLinks`.
+     *
+     * If an object is provided, it may contain a `userAgent` property to specify which user-agent
+     * should be used when checking the robots.txt file. If not provided, the default user-agent `*` will be used.
      */
-    respectRobotsTxtFile?: boolean;
+    respectRobotsTxtFile?: boolean | {
+        userAgent?: string;
+    };
     /**
      * When a request is skipped for some reason, you can use this callback to act on it.
-     * This is currently fired only for requests skipped based on robots.txt file.
+     * This is currently fired for requests skipped
+     * 1. based on robots.txt file,
+     * 2. because they don't match enqueueLinks filters,
+     * 3. because they are redirected to a URL that doesn't match the enqueueLinks strategy,
+     * 4. or because the {@link BasicCrawlerOptions.maxRequestsPerCrawl|`maxRequestsPerCrawl`} limit has been reached
      */
     onSkippedRequest?: SkippedRequestCallback;
     /** @internal */
@@ -223,6 +252,23 @@ export interface BasicCrawlerOptions<Context extends CrawlingContext = BasicCraw
      * Defaults to a new instance of {@link GotScrapingHttpClient}
      */
     httpClient?: BaseHttpClient;
+    /**
+     * If set, the crawler will be configured for all connections to use
+     * the Proxy URLs provided and rotated according to the configuration.
+     */
+    proxyConfiguration?: ProxyConfiguration;
+    /**
+     * A unique identifier for the crawler instance. This ID is used to isolate the state returned by
+     * {@link BasicCrawler.useState|`crawler.useState()`} from other crawler instances.
+     *
+     * When multiple crawler instances use `useState()` without an explicit `id`, they will share the same
+     * state object for backward compatibility. A warning will be logged in this case.
+     *
+     * To ensure each crawler has its own isolated state that also persists across script restarts
+     * (e.g., during Apify migrations), provide a stable, unique ID for each crawler instance.
+     *
+     */
+    id?: string;
 }
 /**
  * A set of options that you can toggle to enable experimental features in Crawlee.
@@ -303,9 +349,14 @@ export interface CrawlerExperiments {
  * ```
  * @category Crawlers
  */
-export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlingContext> {
+export declare class BasicCrawler<Context extends CrawlingContext = CrawlingContext, ContextExtension = Dictionary<never>, ExtendedContext extends Context = Context & ContextExtension> {
     readonly config: Configuration;
     protected static readonly CRAWLEE_STATE_KEY = "CRAWLEE_STATE";
+    /**
+     * Tracks crawler instances that accessed shared state without having an explicit id.
+     * Used to detect and warn about multiple crawlers sharing the same state.
+     */
+    private static useStateCrawlerIds;
     /**
      * A reference to the underlying {@link Statistics} class that collects and logs run statistics for requests.
      */
@@ -321,6 +372,10 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
      * Only available if used by the crawler.
      */
     requestQueue?: RequestProvider;
+    /**
+     * The main request-handling component of the crawler. It's initialized during the crawler startup.
+     */
+    protected requestManager?: IRequestManager;
     /**
      * A reference to the underlying {@link SessionPool} class that manages the crawler's {@link Session|sessions}.
      * Only available if used by the crawler.
@@ -334,40 +389,59 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
      * or to abort it by calling {@link AutoscaledPool.abort|`autoscaledPool.abort()`}.
      */
     autoscaledPool?: AutoscaledPool;
+    /**
+     * A reference to the underlying {@link ProxyConfiguration} class that manages the crawler's proxies.
+     * Only available if used by the crawler.
+     */
+    proxyConfiguration?: ProxyConfiguration;
     /**
      * Default {@link Router} instance that will be used if we don't specify any {@link BasicCrawlerOptions.requestHandler|`requestHandler`}.
      * See {@link Router.addHandler|`router.addHandler()`} and {@link Router.addDefaultHandler|`router.addDefaultHandler()`}.
      */
-    readonly router: RouterHandler<LoadedContext<Context>>;
+    readonly router: RouterHandler<Context>;
+    private contextPipelineBuilder;
+    private _contextPipeline?;
+    get contextPipeline(): ContextPipeline<CrawlingContext, ExtendedContext>;
     running: boolean;
     hasFinishedBefore: boolean;
     readonly log: Log;
-    protected requestHandler: RequestHandler<Context>;
-    protected errorHandler?: ErrorHandler<Context>;
-    protected failedRequestHandler?: ErrorHandler<Context>;
+    protected requestHandler: RequestHandler<ExtendedContext>;
+    protected errorHandler?: ErrorHandler<CrawlingContext, ExtendedContext>;
+    protected failedRequestHandler?: ErrorHandler<CrawlingContext, ExtendedContext>;
     protected requestHandlerTimeoutMillis: number;
     protected internalTimeoutMillis: number;
     protected maxRequestRetries: number;
+    protected maxCrawlDepth?: number;
     protected sameDomainDelayMillis: number;
     protected domainAccessedTime: Map<string, number>;
     protected maxSessionRotations: number;
+    protected maxRequestsPerCrawl?: number;
     protected handledRequestsCount: number;
     protected statusMessageLoggingInterval: number;
     protected statusMessageCallback?: StatusMessageCallback;
     protected sessionPoolOptions: SessionPoolOptions;
     protected useSessionPool: boolean;
-    protected crawlingContexts: Map<string, Context>;
     protected autoscaledPoolOptions: AutoscaledPoolOptions;
     protected events: EventManager;
     protected httpClient: BaseHttpClient;
     protected retryOnBlocked: boolean;
-    protected respectRobotsTxtFile: boolean;
+    protected respectRobotsTxtFile: boolean | {
+        userAgent?: string;
+    };
     protected onSkippedRequest?: SkippedRequestCallback;
     private _closeEvents?;
+    private shouldLogMaxProcessedRequestsExceeded;
+    private shouldLogMaxEnqueuedRequestsExceeded;
     private experiments;
     private readonly robotsTxtFileCache;
     private _experimentWarnings;
+    private readonly crawlerId;
+    private readonly hasExplicitId;
     protected static optionsShape: {
+// @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
@@ -388,12 +462,16 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
         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
@@ -401,7 +479,7 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
 // @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").BooleanPredicate & import("ow").BasePredicate<boolean | undefined>;
+        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
@@ -420,11 +498,14 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
         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>;
+// @ts-ignore optional peer dependency or compatibility with es2022
+        id: import("ow").StringPredicate & import("ow").BasePredicate<string | undefined>;
     };
     /**
      * All `BasicCrawler` parameters are passed via an options object.
      */
-    constructor(options?: BasicCrawlerOptions<Context>, config?: Configuration);
+    constructor(options?: BasicCrawlerOptions<Context, ContextExtension, ExtendedContext> & RequireContextPipeline<CrawlingContext, Context>, // cast because the constructor logic handles missing `contextPipelineBuilder` - the type is just for DX
+    config?: Configuration);
     /**
      * Checks if the given error is a proxy error by comparing its message to a list of known proxy error messages.
      * Used for retrying requests that failed due to proxy errors.
@@ -432,12 +513,6 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
      * @param error The error to check.
      */
     protected isProxyError(error: Error): boolean;
-    /**
-     * Checks whether the given crawling context is getting blocked by anti-bot protection using several heuristics.
-     * Returns `false` if the request is not blocked, otherwise returns a string with a description of the block reason.
-     * @param _crawlingContext The crawling context to check.
-     */
-    protected isRequestBlocked(_crawlingContext: Context): Promise<string | false>;
     /**
      * This method is periodically called by the crawler, every `statusMessageLoggingInterval` seconds.
      */
@@ -453,15 +528,20 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
      * @param [requests] The requests to add.
      * @param [options] Options for the request queue.
      */
-    run(requests?: (string | Request | RequestOptions)[], options?: CrawlerRunOptions): Promise<FinalStatistics>;
+    run(requests?: RequestsLike, options?: CrawlerRunOptions): Promise<FinalStatistics>;
     /**
      * Gracefully stops the current run of the crawler.
      *
      * All the tasks active at the time of calling this method will be allowed to finish.
+     *
+     * To stop the crawler immediately, use {@link BasicCrawler.teardown|`crawler.teardown()`} instead.
      */
     stop(message?: string): void;
     getRequestQueue(): Promise<RequestProvider>;
     useState<State extends Dictionary = Dictionary>(defaultValue?: State): Promise<State>;
+    protected get pendingRequestCountApproximation(): number;
+    protected calculateEnqueuedRequestLimit(explicitLimit?: number): number | undefined;
+    protected handleSkippedRequest(options: Parameters<SkippedRequestCallback>[0]): Promise<void>;
     /**
      * Adds requests to the queue in batches. By default, it will resolve after the initial batch is added, and continue
      * adding the rest in background. You can configure the batch size via `batchSize` option and the sleep time in between
@@ -473,7 +553,7 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
      * @param requests The requests to add
      * @param options Options for the request queue
      */
-    addRequests(requests: (string | Source)[], options?: CrawlerAddRequestsOptions): Promise<CrawlerAddRequestsResult>;
+    addRequests(requests: ReadonlyDeep<RequestsLike>, options?: CrawlerAddRequestsOptions): Promise<CrawlerAddRequestsResult>;
     /**
      * Pushes data to the specified {@link Dataset}, or the default crawler {@link Dataset} by calling {@link Dataset.pushData}.
      */
@@ -491,8 +571,11 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
      * Supported formats are currently 'json' and 'csv', and will be inferred from the `path` automatically.
      */
     exportData<Data>(path: string, format?: 'json' | 'csv', options?: DatasetExportOptions): Promise<Data[]>;
+    /**
+     * Initializes the crawler.
+     */
     protected _init(): Promise<void>;
-    protected _runRequestHandler(crawlingContext: Context): Promise<void>;
+    protected runRequestHandler(crawlingContext: CrawlingContext): Promise<void>;
     /**
      * Handles blocked request
      */
@@ -501,31 +584,42 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
     protected getRobotsTxtFileForUrl(url: string): Promise<RobotsTxtFile | undefined>;
     protected _pauseOnMigration(): Promise<void>;
     /**
-     * Fetches request from either RequestList or RequestQueue. If request comes from a RequestList
-     * and RequestQueue is present then enqueues it to the queue first.
+     * Initializes the RequestManager based on the configured requestList and requestQueue.
      */
-    protected _fetchNextRequest(): Promise<Request<Dictionary> | null | undefined>;
+    private initializeRequestManager;
     /**
-     * Executed when `errorHandler` finishes or the request is successful.
-     * Can be used to clean up orphaned browser pages.
+     * Fetches the next request to process from the underlying request provider.
      */
-    protected _cleanupContext(_crawlingContext: Context): Promise<void>;
+    protected _fetchNextRequest(): Promise<Request<Dictionary> | null>;
     /**
      * Delays processing of the request based on the `sameDomainDelaySecs` option,
      * adding it back to the queue after the timeout passes. Returns `true` if the request
      * should be ignored and will be reclaimed to the queue once ready.
      */
-    protected delayRequest(request: Request, source: IRequestList | RequestProvider): boolean;
+    protected delayRequest(request: Request, source: IRequestList | RequestProvider | IRequestManager): boolean;
     /**
      * Wrapper around requestHandler that fetches requests from RequestList/RequestQueue
      * then retries them in a case of an error, etc.
      */
     protected _runTaskFunction(): Promise<void>;
     /**
-     * Run async callback with given timeout and retry.
+     * Wrapper around the crawling context's `enqueueLinks` method:
+     * - Injects `crawlDepth` to each request being added based on the crawling context request.
+     * - Provides defaults for the `enqueueLinks` options based on the crawler configuration.
+     *      - These options can be overridden by the user.
+     * @internal
+     */
+    protected enqueueLinksWithCrawlDepth(options: SetRequired<EnqueueLinksOptions, 'urls'>, request: Request<Dictionary>, requestQueue: RequestProvider): Promise<BatchAddRequestsResult>;
+    /**
+     * Generator function that yields requests injected with the given crawl depth.
+     * @internal
+     */
+    protected addCrawlDepthRequestGenerator(requests: RequestsLike, newRequestDepth: number): AsyncGenerator<Source, void, undefined>;
+    /**
+     * Run async callback with given timeout and retry. Returns the result of the callback.
      * @ignore
      */
-    protected _timeoutAndRetry(handler: () => Promise<unknown>, timeout: number, error: Error | string, maxRetries?: number, retried?: number): Promise<void>;
+    protected _timeoutAndRetry<T>(handler: () => Promise<T>, timeout: number, error: Error | string, maxRetries?: number, retried?: number): Promise<T>;
     /**
      * Returns true if either RequestList or RequestQueue have a request ready for processing.
      */
@@ -535,12 +629,17 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
      */
     protected _defaultIsFinishedFunction(): Promise<boolean>;
     private _rotateSession;
+    /**
+     * Unwraps errors thrown by the context pipeline to get the actual user error.
+     * RequestHandlerError and ContextPipelineInitializationError wrap the actual error.
+     */
+    private unwrapError;
     /**
      * Handles errors thrown by user provided requestHandler()
      */
-    protected _requestFunctionErrorHandler(error: Error, crawlingContext: Context, source: IRequestList | RequestProvider): Promise<void>;
+    protected _requestFunctionErrorHandler(error: Error, crawlingContext: CrawlingContext, source: IRequestList | IRequestManager): Promise<void>;
     protected _tagUserHandlerError<T>(cb: () => unknown): Promise<T>;
-    protected _handleFailedRequestHandler(crawlingContext: Context, error: Error): Promise<void>;
+    protected _handleFailedRequestHandler(crawlingContext: CrawlingContext, error: Error): Promise<void>;
     /**
      * Resolves the most verbose error message from a thrown error
      * @param error The error received
@@ -549,23 +648,21 @@ export declare class BasicCrawler<Context extends CrawlingContext = BasicCrawlin
     protected _getMessageFromError(error: Error, forceStack?: boolean): string | TimeoutError | undefined;
     protected _canRequestBeRetried(request: Request, error: Error): boolean;
     /**
-     * Updates handledRequestsCount from possibly stored counts,
-     * usually after worker migration. Since one of the stores
-     * needs to have priority when both are present,
-     * it is the request queue, because generally, the request
-     * list will first be dumped into the queue and then left
-     * empty.
+     * Updates handledRequestsCount from possibly stored counts, usually after worker migration.
      */
     protected _loadHandledRequestCount(): Promise<void>;
     protected _executeHooks<HookLike extends (...args: any[]) => Awaitable<void>>(hooks: HookLike[], ...args: Parameters<HookLike>): Promise<void>;
     /**
-     * Function for cleaning up after all request are processed.
-     * @ignore
+     * Stops the crawler immediately.
+     *
+     * This method doesn't wait for currently active requests to finish.
+     *
+     * To stop the crawler gracefully (waiting for all running requests to finish), use {@link BasicCrawler.stop|`crawler.stop()`} instead.
      */
     teardown(): Promise<void>;
     protected _getCookieHeaderFromRequest(request: Request): string;
     private _getRequestQueue;
-    protected requestMatchesEnqueueStrategy(request: Request): boolean;
+    private requestMatchesEnqueueStrategy;
 }
 export interface CreateContextOptions {
     request: Request;

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

@@ -1 +1 @@
-{"version":3,"file":"basic-crawler.d.ts","sourceRoot":"","sources":["../../src/internals/basic-crawler.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EACR,yBAAyB,EACzB,wBAAwB,EACxB,qBAAqB,EACrB,cAAc,EACd,eAAe,EACf,oBAAoB,EACpB,mBAAmB,EACnB,YAAY,EACZ,eAAe,EACf,sBAAsB,EACtB,YAAY,EACZ,aAAa,EACb,SAAS,EACT,OAAO,EACP,cAAc,EACd,yBAAyB,EACzB,aAAa,EACb,YAAY,EACZ,OAAO,EACP,kBAAkB,EAClB,sBAAsB,EACtB,MAAM,EACN,iBAAiB,EACjB,cAAc,EACjB,MAAM,eAAe,CAAC;AACvB,OAAO,EACH,cAAc,EACd,aAAa,EAEb,OAAO,EASP,eAAe,EAOf,WAAW,EACX,UAAU,EAEb,MAAM,eAAe,CAAC;AACvB,OAAO,KAAK,EAAE,SAAS,EAAE,sBAAsB,EAAE,UAAU,EAAE,uBAAuB,EAAE,MAAM,gBAAgB,CAAC;AAC7G,OAAO,EAAE,aAAa,EAAuB,MAAM,gBAAgB,CAAC;AAKpE,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,WAAW,CAAC;AAG7C,OAAO,KAAK,EAAE,GAAG,EAAE,MAAM,YAAY,CAAC;AAEtC,OAAO,EAAuB,YAAY,EAAa,MAAM,gBAAgB,CAAC;AAK9E,MAAM,WAAW,oBAAoB,CAAC,QAAQ,SAAS,UAAU,GAAG,UAAU,CAC1E,SAAQ,eAAe,CAAC,YAAY,EAAE,QAAQ,CAAC;IAC/C;;;;;;;;;;;;;;;;;;;;;;OAsBG;IACH,YAAY,CAAC,OAAO,CAAC,EAAE,WAAW,CAAC,mBAAmB,EAAE,MAAM,CAAC,GAAG,OAAO,CAAC,sBAAsB,CAAC,CAAC;CACrG;AAaD,MAAM,MAAM,cAAc,CACtB,OAAO,SAAS,eAAe,GAAG,aAAa,CAAC,oBAAoB,GAAG,yBAAyB,CAAC,IACjG,CAAC,MAAM,EAAE,aAAa,CAAC,OAAO,CAAC,KAAK,SAAS,CAAC,IAAI,CAAC,CAAC;AAExD,MAAM,MAAM,YAAY,CACpB,OAAO,SAAS,eAAe,GAAG,aAAa,CAAC,oBAAoB,GAAG,yBAAyB,CAAC,IACjG,CAAC,MAAM,EAAE,aAAa,CAAC,OAAO,CAAC,EAAE,KAAK,EAAE,KAAK,KAAK,SAAS,CAAC,IAAI,CAAC,CAAC;AAEtE,MAAM,WAAW,2BAA2B,CACxC,OAAO,SAAS,eAAe,GAAG,oBAAoB,EACtD,OAAO,SAAS,YAAY,CAAC,GAAG,CAAC,GAAG,YAAY,CAAC,OAAO,CAAC;IAEzD,KAAK,EAAE,cAAc,CAAC;IACtB,OAAO,EAAE,OAAO,CAAC;IACjB,aAAa,EAAE,cAAc,CAAC;IAC9B,OAAO,EAAE,MAAM,CAAC;CACnB;AAED,MAAM,MAAM,qBAAqB,CAC7B,OAAO,SAAS,eAAe,GAAG,oBAAoB,EACtD,OAAO,SAAS,YAAY,CAAC,GAAG,CAAC,GAAG,YAAY,CAAC,OAAO,CAAC,IACzD,CAAC,MAAM,EAAE,2BAA2B,CAAC,OAAO,EAAE,OAAO,CAAC,KAAK,SAAS,CAAC,IAAI,CAAC,CAAC;AAE/E,MAAM,WAAW,mBAAmB,CAAC,OAAO,SAAS,eAAe,GAAG,oBAAoB;IACvF;;;;;;;;;;;;;;;;OAgBG;IACH,cAAc,CAAC,EAAE,cAAc,CAAC,aAAa,CAAC,OAAO,CAAC,CAAC,CAAC;IAExD;;;;;OAKG;IACH,WAAW,CAAC,EAAE,YAAY,CAAC;IAE3B;;;;;OAKG;IACH,YAAY,CAAC,EAAE,eAAe,CAAC;IAE/B;;;OAGG;IACH,yBAAyB,CAAC,EAAE,MAAM,CAAC;IAEnC;;;;;;;;OAQG;IACH,YAAY,CAAC,EAAE,YAAY,CAAC,OAAO,CAAC,CAAC;IAErC;;;;;;;OAOG;IACH,oBAAoB,CAAC,EAAE,YAAY,CAAC,OAAO,CAAC,CAAC;IAE7C;;;;;;;;OAQG;IACH,iBAAiB,CAAC,EAAE,MAAM,CAAC;IAE3B;;;OAGG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAE7B;;;;;;OAMG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAE7B;;;;OAIG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAE7B;;;;;;OAMG;IACH,qBAAqB,CAAC,EAAE,qBAAqB,CAAC;IAE9C;;;;;OAKG;IACH,cAAc,CAAC,EAAE,MAAM,CAAC;IAExB;;;OAGG;IACH,cAAc,CAAC,EAAE,MAAM,CAAC;IAExB;;;;OAIG;IACH,oBAAoB,CAAC,EAAE,MAAM,CAAC;IAE9B;;;;OAIG;IACH,SAAS,CAAC,EAAE,OAAO,CAAC;IAEpB;;;OAGG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;IAEzB;;OAEG;IACH,kBAAkB,CAAC,EAAE,kBAAkB,CAAC;IAExC;;OAEG;IACH,4BAA4B,CAAC,EAAE,MAAM,CAAC;IAEtC;;;;;;;;;;;;;;;OAeG;IACH,qBAAqB,CAAC,EAAE,qBAAqB,CAAC;IAE9C;;;;;;OAMG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;IAEzB;;;OAGG;IACH,oBAAoB,CAAC,EAAE,OAAO,CAAC;IAE/B;;;OAGG;IACH,gBAAgB,CAAC,EAAE,sBAAsB,CAAC;IAE1C,gBAAgB;IAChB,GAAG,CAAC,EAAE,GAAG,CAAC;IAEV;;;OAGG;IACH,WAAW,CAAC,EAAE,kBAAkB,CAAC;IAEjC;;;OAGG;IACH,iBAAiB,CAAC,EAAE,iBAAiB,CAAC;IAEtC;;;OAGG;IACH,UAAU,CAAC,EAAE,cAAc,CAAC;CAC/B;AAED;;;;;GAKG;AACH,MAAM,WAAW,kBAAkB;IAC/B;;;;;OAKG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;CAC5B;AAED;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA+DG;AACH,qBAAa,YAAY,CAAC,OAAO,SAAS,eAAe,GAAG,oBAAoB;IAqHxE,QAAQ,CAAC,MAAM;IApHnB,SAAS,CAAC,MAAM,CAAC,QAAQ,CAAC,iBAAiB,mBAAmB;IAE9D;;OAEG;IACH,QAAQ,CAAC,KAAK,EAAE,UAAU,CAAC;IAE3B;;;OAGG;IACH,WAAW,CAAC,EAAE,YAAY,CAAC;IAE3B;;;;OAIG;IACH,YAAY,CAAC,EAAE,eAAe,CAAC;IAE/B;;;OAGG;IACH,WAAW,CAAC,EAAE,WAAW,CAAC;IAE1B;;;;;;OAMG;IACH,cAAc,CAAC,EAAE,cAAc,CAAC;IAEhC;;;OAGG;IACH,QAAQ,CAAC,MAAM,EAAE,aAAa,CAAC,aAAa,CAAC,OAAO,CAAC,CAAC,CAA2C;IAEjG,OAAO,UAAS;IAChB,iBAAiB,UAAS;IAE1B,QAAQ,CAAC,GAAG,EAAE,GAAG,CAAC;IAClB,SAAS,CAAC,cAAc,EAAG,cAAc,CAAC,OAAO,CAAC,CAAC;IACnD,SAAS,CAAC,YAAY,CAAC,EAAE,YAAY,CAAC,OAAO,CAAC,CAAC;IAC/C,SAAS,CAAC,oBAAoB,CAAC,EAAE,YAAY,CAAC,OAAO,CAAC,CAAC;IACvD,SAAS,CAAC,2BAA2B,EAAG,MAAM,CAAC;IAC/C,SAAS,CAAC,qBAAqB,EAAE,MAAM,CAAC;IACxC,SAAS,CAAC,iBAAiB,EAAE,MAAM,CAAC;IACpC,SAAS,CAAC,qBAAqB,EAAE,MAAM,CAAC;IACxC,SAAS,CAAC,kBAAkB,EAAE,GAAG,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAClD,SAAS,CAAC,mBAAmB,EAAE,MAAM,CAAC;IACtC,SAAS,CAAC,oBAAoB,EAAE,MAAM,CAAC;IACvC,SAAS,CAAC,4BAA4B,EAAE,MAAM,CAAC;IAC/C,SAAS,CAAC,qBAAqB,CAAC,EAAE,qBAAqB,CAAC;IACxD,SAAS,CAAC,kBAAkB,EAAE,kBAAkB,CAAC;IACjD,SAAS,CAAC,cAAc,EAAE,OAAO,CAAC;IAClC,SAAS,CAAC,gBAAgB,uBAA8B;IACxD,SAAS,CAAC,qBAAqB,EAAE,qBAAqB,CAAC;IACvD,SAAS,CAAC,MAAM,EAAE,YAAY,CAAC;IAC/B,SAAS,CAAC,UAAU,EAAE,cAAc,CAAC;IACrC,SAAS,CAAC,cAAc,EAAE,OAAO,CAAC;IAClC,SAAS,CAAC,oBAAoB,EAAE,OAAO,CAAC;IACxC,SAAS,CAAC,gBAAgB,CAAC,EAAE,sBAAsB,CAAC;IACpD,OAAO,CAAC,YAAY,CAAC,CAAU;IAE/B,OAAO,CAAC,WAAW,CAAqB;IACxC,OAAO,CAAC,QAAQ,CAAC,kBAAkB,CAA0B;IAC7D,OAAO,CAAC,mBAAmB,CAA0D;IAErF,SAAS,CAAC,MAAM,CAAC,YAAY;;;;;;;;;;;;;;;;;;;;;;;;;;;MAqC3B;IAEF;;OAEG;gBAEC,OAAO,GAAE,mBAAmB,CAAC,OAAO,CAAM,EACjC,MAAM,gBAAkC;IA6KrD;;;;;OAKG;IACH,SAAS,CAAC,YAAY,CAAC,KAAK,EAAE,KAAK,GAAG,OAAO;IAI7C;;;;OAIG;cACa,gBAAgB,CAAC,gBAAgB,EAAE,OAAO,GAAG,OAAO,CAAC,MAAM,GAAG,KAAK,CAAC;IAIpF;;OAEG;IACG,gBAAgB,CAAC,OAAO,EAAE,MAAM,EAAE,OAAO,GAAE,uBAA4B;IAmB7E,OAAO,CAAC,iBAAiB;IAgDzB;;;;;;;;;OASG;IACG,GAAG,CAAC,QAAQ,CAAC,EAAE,CAAC,MAAM,GAAG,OAAO,GAAG,cAAc,CAAC,EAAE,EAAE,OAAO,CAAC,EAAE,iBAAiB,GAAG,OAAO,CAAC,eAAe,CAAC;IA8GlH;;;;OAIG;IACH,IAAI,CAAC,OAAO,SAA6C,GAAG,IAAI;IAY1D,eAAe;IAYf,QAAQ,CAAC,KAAK,SAAS,UAAU,GAAG,UAAU,EAAE,YAAY,GAAS,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC;IAKjG;;;;;;;;;;OAUG;IACG,WAAW,CACb,QAAQ,EAAE,CAAC,MAAM,GAAG,MAAM,CAAC,EAAE,EAC7B,OAAO,GAAE,yBAA8B,GACxC,OAAO,CAAC,wBAAwB,CAAC;IAsCpC;;OAEG;IACG,QAAQ,CAAC,IAAI,EAAE,UAAU,CAAC,OAAO,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,eAAe,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC;IAKjG;;OAEG;IACG,UAAU,CAAC,QAAQ,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC,OAAO,CAAC;IAIrD;;OAEG;IACG,OAAO,CAAC,GAAG,IAAI,EAAE,UAAU,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,GAAG,UAAU,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC;IAKtF;;;OAGG;IACG,UAAU,CAAC,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,MAAM,GAAG,KAAK,EAAE,OAAO,CAAC,EAAE,oBAAoB,GAAG,OAAO,CAAC,IAAI,EAAE,CAAC;cAoC9F,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;cAoBtB,kBAAkB,CAAC,eAAe,EAAE,OAAO,GAAG,OAAO,CAAC,IAAI,CAAC;IAI3E;;OAEG;IACH,SAAS,CAAC,sBAAsB,CAAC,OAAO,EAAE,OAAO,EAAE,UAAU,EAAE,MAAM;YAQvD,6BAA6B;cAS3B,sBAAsB,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC,aAAa,GAAG,SAAS,CAAC;cAuBvE,iBAAiB;IAuCjC;;;OAGG;cACa,iBAAiB;IAyBjC;;;OAGG;cACa,eAAe,CAAC,gBAAgB,EAAE,OAAO;IAEzD;;;;OAIG;IACH,SAAS,CAAC,YAAY,CAAC,OAAO,EAAE,OAAO,EAAE,MAAM,EAAE,YAAY,GAAG,eAAe;IAsC/E;;;OAGG;cACa,gBAAgB;IA0JhC;;;OAGG;cACa,gBAAgB,CAC5B,OAAO,EAAE,MAAM,OAAO,CAAC,OAAO,CAAC,EAC/B,OAAO,EAAE,MAAM,EACf,KAAK,EAAE,KAAK,GAAG,MAAM,EACrB,UAAU,SAAI,EACd,OAAO,SAAI,GACZ,OAAO,CAAC,IAAI,CAAC;IAehB;;OAEG;cACa,oBAAoB;IASpC;;OAEG;cACa,0BAA0B;YAS5B,cAAc;IAQ5B;;OAEG;cACa,4BAA4B,CACxC,KAAK,EAAE,KAAK,EACZ,eAAe,EAAE,OAAO,EACxB,MAAM,EAAE,YAAY,GAAG,eAAe,GACvC,OAAO,CAAC,IAAI,CAAC;cAyDA,oBAAoB,CAAC,CAAC,EAAE,EAAE,EAAE,MAAM,OAAO,GAAG,OAAO,CAAC,CAAC,CAAC;cAStD,2BAA2B,CAAC,eAAe,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,GAAG,OAAO,CAAC,IAAI,CAAC;IAYlG;;;;OAIG;IACH,SAAS,CAAC,oBAAoB,CAAC,KAAK,EAAE,KAAK,EAAE,UAAU,UAAQ;IAmB/D,SAAS,CAAC,oBAAoB,CAAC,OAAO,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK;IAoB7D;;;;;;;OAOG;cACa,wBAAwB,IAAI,OAAO,CAAC,IAAI,CAAC;cAQzC,aAAa,CAAC,QAAQ,SAAS,CAAC,GAAG,IAAI,EAAE,GAAG,EAAE,KAAK,SAAS,CAAC,IAAI,CAAC,EAC9E,KAAK,EAAE,QAAQ,EAAE,EACjB,GAAG,IAAI,EAAE,UAAU,CAAC,QAAQ,CAAC;IASjC;;;OAGG;IACG,QAAQ,IAAI,OAAO,CAAC,IAAI,CAAC;IAc/B,SAAS,CAAC,2BAA2B,CAAC,OAAO,EAAE,OAAO;YAWxC,gBAAgB;IAc9B,SAAS,CAAC,6BAA6B,CAAC,OAAO,EAAE,OAAO;CA6C3D;AAED,MAAM,WAAW,oBAAoB;IACjC,OAAO,EAAE,OAAO,CAAC;IACjB,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,SAAS,CAAC,EAAE,SAAS,CAAC;CACzB;AAED,MAAM,WAAW,yBAA0B,SAAQ,yBAAyB;CAAG;AAE/E,MAAM,WAAW,wBAAyB,SAAQ,wBAAwB;CAAG;AAE7E,MAAM,WAAW,iBAAkB,SAAQ,yBAAyB;IAChE;;;;OAIG;IACH,iBAAiB,CAAC,EAAE,OAAO,CAAC;CAC/B;AAED;;;;;;;;;;;;;;;;;;;;;;;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,0BAEzC"}
+{"version":3,"file":"basic-crawler.d.ts","sourceRoot":"","sources":["../../src/internals/basic-crawler.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EACR,yBAAyB,EACzB,wBAAwB,EACxB,qBAAqB,EACrB,eAAe,EACf,oBAAoB,EACpB,mBAAmB,EACnB,YAAY,EACZ,eAAe,EACf,sBAAsB,EACtB,YAAY,EACZ,eAAe,EACf,kBAAkB,EAClB,OAAO,EACP,YAAY,EAEZ,aAAa,EACb,YAAY,EACZ,OAAO,EACP,kBAAkB,EAClB,sBAAsB,EACtB,MAAM,EACN,iBAAiB,EACjB,cAAc,EACjB,MAAM,eAAe,CAAC;AACvB,OAAO,EACH,cAAc,EACd,aAAa,EACb,eAAe,EAKf,OAAO,EAYP,eAAe,EAOf,WAAW,EACX,UAAU,EAEb,MAAM,eAAe,CAAC;AACvB,OAAO,KAAK,EACR,SAAS,EACT,cAAc,EACd,sBAAsB,EACtB,UAAU,EACV,SAAS,EACT,uBAAuB,EAC1B,MAAM,gBAAgB,CAAC;AACxB,OAAO,EAA8C,aAAa,EAAuB,MAAM,gBAAgB,CAAC;AAKhH,OAAO,KAAK,EAAE,YAAY,EAAE,WAAW,EAAE,MAAM,WAAW,CAAC;AAG3D,OAAO,KAAK,EAAE,GAAG,EAAE,MAAM,YAAY,CAAC;AAEtC,OAAO,EAAuB,YAAY,EAAa,MAAM,gBAAgB,CAAC;AAK9E,MAAM,WAAW,oBAAoB,CAAC,QAAQ,SAAS,UAAU,GAAG,UAAU,CAAE,SAAQ,eAAe,CAAC,QAAQ,CAAC;CAAG;AAapH,MAAM,MAAM,cAAc,CAAC,OAAO,SAAS,eAAe,GAAG,eAAe,IAAI,CAAC,MAAM,EAAE,OAAO,KAAK,SAAS,CAAC,IAAI,CAAC,CAAC;AAErH,MAAM,MAAM,YAAY,CACpB,OAAO,SAAS,eAAe,GAAG,eAAe,EACjD,eAAe,SAAS,OAAO,GAAG,OAAO,IACzC,CAAC,MAAM,EAAE,OAAO,GAAG,OAAO,CAAC,eAAe,CAAC,EAAE,KAAK,EAAE,KAAK,KAAK,SAAS,CAAC,IAAI,CAAC,CAAC;AAElF,MAAM,WAAW,2BAA2B,CACxC,OAAO,SAAS,eAAe,GAAG,oBAAoB,EACtD,OAAO,SAAS,YAAY,CAAC,GAAG,CAAC,GAAG,YAAY,CAAC,OAAO,CAAC;IAEzD,KAAK,EAAE,cAAc,CAAC;IACtB,OAAO,EAAE,OAAO,CAAC;IACjB,aAAa,EAAE,cAAc,CAAC;IAC9B,OAAO,EAAE,MAAM,CAAC;CACnB;AAED,MAAM,MAAM,qBAAqB,CAC7B,OAAO,SAAS,eAAe,GAAG,oBAAoB,EACtD,OAAO,SAAS,YAAY,CAAC,GAAG,CAAC,GAAG,YAAY,CAAC,OAAO,CAAC,IACzD,CAAC,MAAM,EAAE,2BAA2B,CAAC,OAAO,EAAE,OAAO,CAAC,KAAK,SAAS,CAAC,IAAI,CAAC,CAAC;AAE/E,MAAM,MAAM,sBAAsB,CAC9B,kBAAkB,SAAS,eAAe,EAC1C,gBAAgB,SAAS,kBAAkB,IAC3C,kBAAkB,SAAS,gBAAgB,GACzC,EAAE,GACF;IAAE,sBAAsB,EAAE,MAAM,eAAe,CAAC,eAAe,EAAE,gBAAgB,CAAC,CAAA;CAAE,CAAC;AAE3F,MAAM,WAAW,mBAAmB,CAChC,OAAO,SAAS,eAAe,GAAG,eAAe,EACjD,gBAAgB,GAAG,UAAU,CAAC,KAAK,CAAC,EACpC,eAAe,SAAS,OAAO,GAAG,OAAO,GAAG,gBAAgB;IAE5D;;;;;;;;;;;;;;;;OAgBG;IACH,cAAc,CAAC,EAAE,cAAc,CAAC,eAAe,CAAC,CAAC;IAEjD;;;;;;;;;;;;;;;;;;;;OAoBG;IACH,aAAa,CAAC,EAAE,CAAC,OAAO,EAAE,OAAO,KAAK,SAAS,CAAC,gBAAgB,CAAC,CAAC;IAElE;;;;OAIG;IACH,sBAAsB,CAAC,EAAE,MAAM,eAAe,CAAC,eAAe,EAAE,OAAO,CAAC,CAAC;IAEzE;;;;;OAKG;IACH,WAAW,CAAC,EAAE,YAAY,CAAC;IAE3B;;;;;OAKG;IACH,YAAY,CAAC,EAAE,eAAe,CAAC;IAE/B;;;;;OAKG;IACH,cAAc,CAAC,EAAE,eAAe,CAAC;IAEjC;;;OAGG;IACH,yBAAyB,CAAC,EAAE,MAAM,CAAC;IAEnC;;;;;;;;OAQG;IACH,YAAY,CAAC,EAAE,YAAY,CAAC,eAAe,EAAE,eAAe,CAAC,CAAC;IAE9D;;;;;;;OAOG;IACH,oBAAoB,CAAC,EAAE,YAAY,CAAC,eAAe,EAAE,eAAe,CAAC,CAAC;IAEtE;;;;;;;;OAQG;IACH,iBAAiB,CAAC,EAAE,MAAM,CAAC;IAE3B;;;OAGG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAE7B;;;;;;OAMG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAE7B;;;;OAIG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAE7B;;;;OAIG;IACH,aAAa,CAAC,EAAE,MAAM,CAAC;IAEvB;;;;;;OAMG;IACH,qBAAqB,CAAC,EAAE,qBAAqB,CAAC;IAE9C;;;;;OAKG;IACH,cAAc,CAAC,EAAE,MAAM,CAAC;IAExB;;;OAGG;IACH,cAAc,CAAC,EAAE,MAAM,CAAC;IAExB;;;;OAIG;IACH,oBAAoB,CAAC,EAAE,MAAM,CAAC;IAE9B;;;;OAIG;IACH,SAAS,CAAC,EAAE,OAAO,CAAC;IAEpB;;;OAGG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;IAEzB;;OAEG;IACH,kBAAkB,CAAC,EAAE,kBAAkB,CAAC;IAExC;;OAEG;IACH,4BAA4B,CAAC,EAAE,MAAM,CAAC;IAEtC;;;;;;;;;;;;;;;OAeG;IACH,qBAAqB,CAAC,EAAE,qBAAqB,CAAC;IAE9C;;;;;;OAMG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;IAEzB;;;;;;OAMG;IACH,oBAAoB,CAAC,EAAE,OAAO,GAAG;QAAE,SAAS,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;IAExD;;;;;;;OAOG;IACH,gBAAgB,CAAC,EAAE,sBAAsB,CAAC;IAE1C,gBAAgB;IAChB,GAAG,CAAC,EAAE,GAAG,CAAC;IAEV;;;OAGG;IACH,WAAW,CAAC,EAAE,kBAAkB,CAAC;IAEjC;;;OAGG;IACH,iBAAiB,CAAC,EAAE,iBAAiB,CAAC;IAEtC;;;OAGG;IACH,UAAU,CAAC,EAAE,cAAc,CAAC;IAE5B;;;OAGG;IACH,kBAAkB,CAAC,EAAE,kBAAkB,CAAC;IAExC;;;;;;;;;;OAUG;IACH,EAAE,CAAC,EAAE,MAAM,CAAC;CACf;AAED;;;;;GAKG;AACH,MAAM,WAAW,kBAAkB;IAC/B;;;;;OAKG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;CAC5B;AAED;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA+DG;AACH,qBAAa,YAAY,CACrB,OAAO,SAAS,eAAe,GAAG,eAAe,EACjD,gBAAgB,GAAG,UAAU,CAAC,KAAK,CAAC,EACpC,eAAe,SAAS,OAAO,GAAG,OAAO,GAAG,gBAAgB;IA8JxD,QAAQ,CAAC,MAAM;IA5JnB,SAAS,CAAC,MAAM,CAAC,QAAQ,CAAC,iBAAiB,mBAAmB;IAE9D;;;OAGG;IACH,OAAO,CAAC,MAAM,CAAC,kBAAkB,CAAqB;IAEtD;;OAEG;IACH,QAAQ,CAAC,KAAK,EAAE,UAAU,CAAC;IAE3B;;;OAGG;IACH,WAAW,CAAC,EAAE,YAAY,CAAC;IAE3B;;;;OAIG;IACH,YAAY,CAAC,EAAE,eAAe,CAAC;IAE/B;;OAEG;IACH,SAAS,CAAC,cAAc,CAAC,EAAE,eAAe,CAAC;IAE3C;;;OAGG;IACH,WAAW,CAAC,EAAE,WAAW,CAAC;IAE1B;;;;;;OAMG;IACH,cAAc,CAAC,EAAE,cAAc,CAAC;IAEhC;;;OAGG;IACH,kBAAkB,CAAC,EAAE,kBAAkB,CAAC;IAExC;;;OAGG;IACH,QAAQ,CAAC,MAAM,EAAE,aAAa,CAAC,OAAO,CAAC,CAA4B;IAEnE,OAAO,CAAC,sBAAsB,CAA0D;IACxF,OAAO,CAAC,gBAAgB,CAAC,CAAoD;IAE7E,IAAI,eAAe,IAAI,eAAe,CAAC,eAAe,EAAE,eAAe,CAAC,CAMvE;IAED,OAAO,UAAS;IAChB,iBAAiB,UAAS;IAE1B,QAAQ,CAAC,GAAG,EAAE,GAAG,CAAC;IAClB,SAAS,CAAC,cAAc,EAAG,cAAc,CAAC,eAAe,CAAC,CAAC;IAC3D,SAAS,CAAC,YAAY,CAAC,EAAE,YAAY,CAAC,eAAe,EAAE,eAAe,CAAC,CAAC;IACxE,SAAS,CAAC,oBAAoB,CAAC,EAAE,YAAY,CAAC,eAAe,EAAE,eAAe,CAAC,CAAC;IAChF,SAAS,CAAC,2BAA2B,EAAG,MAAM,CAAC;IAC/C,SAAS,CAAC,qBAAqB,EAAE,MAAM,CAAC;IACxC,SAAS,CAAC,iBAAiB,EAAE,MAAM,CAAC;IACpC,SAAS,CAAC,aAAa,CAAC,EAAE,MAAM,CAAC;IACjC,SAAS,CAAC,qBAAqB,EAAE,MAAM,CAAC;IACxC,SAAS,CAAC,kBAAkB,EAAE,GAAG,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAClD,SAAS,CAAC,mBAAmB,EAAE,MAAM,CAAC;IACtC,SAAS,CAAC,mBAAmB,CAAC,EAAE,MAAM,CAAC;IACvC,SAAS,CAAC,oBAAoB,SAAK;IACnC,SAAS,CAAC,4BAA4B,EAAE,MAAM,CAAC;IAC/C,SAAS,CAAC,qBAAqB,CAAC,EAAE,qBAAqB,CAAC;IACxD,SAAS,CAAC,kBAAkB,EAAE,kBAAkB,CAAC;IACjD,SAAS,CAAC,cAAc,EAAE,OAAO,CAAC;IAClC,SAAS,CAAC,qBAAqB,EAAE,qBAAqB,CAAC;IACvD,SAAS,CAAC,MAAM,EAAE,YAAY,CAAC;IAC/B,SAAS,CAAC,UAAU,EAAE,cAAc,CAAC;IACrC,SAAS,CAAC,cAAc,EAAE,OAAO,CAAC;IAClC,SAAS,CAAC,oBAAoB,EAAE,OAAO,GAAG;QAAE,SAAS,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;IACjE,SAAS,CAAC,gBAAgB,CAAC,EAAE,sBAAsB,CAAC;IACpD,OAAO,CAAC,YAAY,CAAC,CAAU;IAC/B,OAAO,CAAC,qCAAqC,CAAQ;IACrD,OAAO,CAAC,oCAAoC,CAAQ;IACpD,OAAO,CAAC,WAAW,CAAqB;IACxC,OAAO,CAAC,QAAQ,CAAC,kBAAkB,CAA0B;IAC7D,OAAO,CAAC,mBAAmB,CAA0D;IACrF,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAS;IACnC,OAAO,CAAC,QAAQ,CAAC,aAAa,CAAU;IAExC,SAAS,CAAC,MAAM,CAAC,YAAY;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;MA4C3B;IAEF;;OAEG;gBAEC,OAAO,GAAE,mBAAmB,CAAC,OAAO,EAAE,gBAAgB,EAAE,eAAe,CAAC,GACpE,sBAAsB,CAAC,eAAe,EAAE,OAAO,CAAa,EAAE,wGAAwG;IACjK,MAAM,gBAAkC;IAuOrD;;;;;OAKG;IACH,SAAS,CAAC,YAAY,CAAC,KAAK,EAAE,KAAK,GAAG,OAAO;IAI7C;;OAEG;IACG,gBAAgB,CAAC,OAAO,EAAE,MAAM,EAAE,OAAO,GAAE,uBAA4B;IAmB7E,OAAO,CAAC,iBAAiB;IAgDzB;;;;;;;;;OASG;IACG,GAAG,CAAC,QAAQ,CAAC,EAAE,YAAY,EAAE,OAAO,CAAC,EAAE,iBAAiB,GAAG,OAAO,CAAC,eAAe,CAAC;IA0HzF;;;;;;OAMG;IACH,IAAI,CAAC,OAAO,SAA6C,GAAG,IAAI;IAY1D,eAAe,IAAI,OAAO,CAAC,eAAe,CAAC;IAsB3C,QAAQ,CAAC,KAAK,SAAS,UAAU,GAAG,UAAU,EAAE,YAAY,GAAS,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC;IAsBjG,SAAS,KAAK,gCAAgC,IAAI,MAAM,CAEvD;IAED,SAAS,CAAC,6BAA6B,CAAC,aAAa,CAAC,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS;cAanE,oBAAoB,CAAC,OAAO,EAAE,UAAU,CAAC,sBAAsB,CAAC,CAAC,CAAC,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC;IAmBnG;;;;;;;;;;OAUG;IACG,WAAW,CACb,QAAQ,EAAE,YAAY,CAAC,YAAY,CAAC,EACpC,OAAO,GAAE,yBAA8B,GACxC,OAAO,CAAC,wBAAwB,CAAC;IA4EpC;;OAEG;IACG,QAAQ,CAAC,IAAI,EAAE,UAAU,CAAC,OAAO,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,eAAe,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC;IAKjG;;OAEG;IACG,UAAU,CAAC,QAAQ,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC,OAAO,CAAC;IAIrD;;OAEG;IACG,OAAO,CAAC,GAAG,IAAI,EAAE,UAAU,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,GAAG,UAAU,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC;IAKtF;;;OAGG;IACG,UAAU,CAAC,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,MAAM,GAAG,KAAK,EAAE,OAAO,CAAC,EAAE,oBAAoB,GAAG,OAAO,CAAC,IAAI,EAAE,CAAC;IAmD9G;;OAEG;cACa,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;cAqBtB,iBAAiB,CAAC,eAAe,EAAE,eAAe,GAAG,OAAO,CAAC,IAAI,CAAC;IAUlF;;OAEG;IACH,SAAS,CAAC,sBAAsB,CAAC,OAAO,EAAE,OAAO,EAAE,UAAU,EAAE,MAAM;YAQvD,6BAA6B;cAW3B,sBAAsB,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC,aAAa,GAAG,SAAS,CAAC;cAuBvE,iBAAiB;IAuCjC;;OAEG;YACW,wBAAwB;IAoBtC;;OAEG;cACa,iBAAiB;IAQjC;;;;OAIG;IACH,SAAS,CAAC,YAAY,CAAC,OAAO,EAAE,OAAO,EAAE,MAAM,EAAE,YAAY,GAAG,eAAe,GAAG,eAAe;IAsCjG;;;OAGG;cACa,gBAAgB;IA0JhC;;;;;;OAMG;cACa,0BAA0B,CACtC,OAAO,EAAE,WAAW,CAAC,mBAAmB,EAAE,MAAM,CAAC,EACjD,OAAO,EAAE,OAAO,CAAC,UAAU,CAAC,EAC5B,YAAY,EAAE,eAAe,GAC9B,OAAO,CAAC,sBAAsB,CAAC;IA0BlC;;;OAGG;cACc,6BAA6B,CAC1C,QAAQ,EAAE,YAAY,EACtB,eAAe,EAAE,MAAM,GACxB,cAAc,CAAC,MAAM,EAAE,IAAI,EAAE,SAAS,CAAC;IAW1C;;;OAGG;cACa,gBAAgB,CAAC,CAAC,EAC9B,OAAO,EAAE,MAAM,OAAO,CAAC,CAAC,CAAC,EACzB,OAAO,EAAE,MAAM,EACf,KAAK,EAAE,KAAK,GAAG,MAAM,EACrB,UAAU,SAAI,EACd,OAAO,SAAI,GACZ,OAAO,CAAC,CAAC,CAAC;IAcb;;OAEG;cACa,oBAAoB;IAIpC;;OAEG;cACa,0BAA0B;YAI5B,cAAc;IAQ5B;;;OAGG;IACH,OAAO,CAAC,WAAW;IAWnB;;OAEG;cACa,4BAA4B,CACxC,KAAK,EAAE,KAAK,EACZ,eAAe,EAAE,eAAe,EAChC,MAAM,EAAE,YAAY,GAAG,eAAe,GACvC,OAAO,CAAC,IAAI,CAAC;cA4DA,oBAAoB,CAAC,CAAC,EAAE,EAAE,EAAE,MAAM,OAAO,GAAG,OAAO,CAAC,CAAC,CAAC;cAStD,2BAA2B,CAAC,eAAe,EAAE,eAAe,EAAE,KAAK,EAAE,KAAK,GAAG,OAAO,CAAC,IAAI,CAAC;IAe1G;;;;OAIG;IACH,SAAS,CAAC,oBAAoB,CAAC,KAAK,EAAE,KAAK,EAAE,UAAU,UAAQ;IAmB/D,SAAS,CAAC,oBAAoB,CAAC,OAAO,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK;IAoB7D;;OAEG;cACa,wBAAwB,IAAI,OAAO,CAAC,IAAI,CAAC;cAMzC,aAAa,CAAC,QAAQ,SAAS,CAAC,GAAG,IAAI,EAAE,GAAG,EAAE,KAAK,SAAS,CAAC,IAAI,CAAC,EAC9E,KAAK,EAAE,QAAQ,EAAE,EACjB,GAAG,IAAI,EAAE,UAAU,CAAC,QAAQ,CAAC;IASjC;;;;;;OAMG;IACG,QAAQ,IAAI,OAAO,CAAC,IAAI,CAAC;IAY/B,SAAS,CAAC,2BAA2B,CAAC,OAAO,EAAE,OAAO;YAWxC,gBAAgB;IAc9B,OAAO,CAAC,6BAA6B;CA6CxC;AAED,MAAM,WAAW,oBAAoB;IACjC,OAAO,EAAE,OAAO,CAAC;IACjB,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,SAAS,CAAC,EAAE,SAAS,CAAC;CACzB;AAED,MAAM,WAAW,yBAA0B,SAAQ,yBAAyB;CAAG;AAE/E,MAAM,WAAW,wBAAyB,SAAQ,wBAAwB;CAAG;AAE7E,MAAM,WAAW,iBAAkB,SAAQ,yBAAyB;IAChE;;;;OAIG;IACH,iBAAiB,CAAC,EAAE,OAAO,CAAC;CAC/B;AAED;;;;;;;;;;;;;;;;;;;;;;;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,0BAEzC"}