@crawlee/utils 4.0.0-beta.11 → 4.0.0-beta.110

package/internals/robots.js

@@ -1,7 +1,6 @@
-import { gotScraping } from 'got-scraping';
+import { FetchHttpClient } from '@crawlee/http-client';
 import robotsParser from 'robots-parser';
 import { Sitemap } from './sitemap.js';
-let HTTPError;
 /**
  * Loads and queries information from a [robots.txt file](https://en.wikipedia.org/wiki/Robots.txt).
  *
@@ -21,22 +20,27 @@ let HTTPError;
  * ```
  */
 export class RobotsTxtFile {
-    robots;
-    proxyUrl;
-    constructor(robots, proxyUrl) {
-        this.robots = robots;
-        this.proxyUrl = proxyUrl;
+    #robots;
+    #proxyUrl;
+    #logger;
+    constructor(robots, proxyUrl, logger) {
+        this.#robots = robots;
+        this.#proxyUrl = proxyUrl;
+        this.#logger = logger;
     }
     /**
      * Determine the location of a robots.txt file for a URL and fetch it.
      * @param url the URL to fetch robots.txt for
-     * @param [proxyUrl] a proxy to be used for fetching the robots.txt file
+     * @param [options] additional options
+     * @param [options.signal] an AbortSignal to cancel the request
+     * @param [options.timeoutMillis] timeout in milliseconds for the request
+     * @param [options.proxyUrl] a proxy to be used for fetching the robots.txt file
      */
-    static async find(url, proxyUrl) {
+    static async find(url, options) {
         const robotsTxtFileUrl = new URL(url);
         robotsTxtFileUrl.pathname = '/robots.txt';
         robotsTxtFileUrl.search = '';
-        return RobotsTxtFile.load(robotsTxtFileUrl.toString(), proxyUrl);
+        return RobotsTxtFile.load(robotsTxtFileUrl.toString(), options);
     }
     /**
      * Allows providing the URL and robots.txt content explicitly instead of loading it from the target site.
@@ -48,33 +52,28 @@ export class RobotsTxtFile {
         // @ts-ignore
         return new RobotsTxtFile(robotsParser(url, content), proxyUrl);
     }
-    static async load(url, proxyUrl) {
-        if (!HTTPError) {
-            HTTPError = (await import('got-scraping')).HTTPError;
-        }
-        try {
-            const response = await gotScraping({
-                url,
-                proxyUrl,
-                method: 'GET',
-                responseType: 'text',
-            });
-            // @ts-ignore
-            return new RobotsTxtFile(robotsParser(url.toString(), response.body), proxyUrl);
+    static async load(url, options) {
+        const { proxyUrl, logger, httpClient = new FetchHttpClient() } = options || {};
+        const response = await httpClient.sendRequest(new Request(url, { method: 'GET' }), {
+            proxyUrl,
+            timeoutMillis: options?.timeoutMillis,
+            signal: options?.signal,
+        });
+        if (response.status < 200 || response.status >= 300) {
+            throw new Error(`Failed to load robots.txt from ${url}: HTTP ${response.status}`);
         }
-        catch (e) {
-            if (e instanceof HTTPError && e.response.statusCode === 404) {
-                return new RobotsTxtFile({
-                    isAllowed() {
-                        return true;
-                    },
-                    getSitemaps() {
-                        return [];
-                    },
-                }, proxyUrl);
-            }
-            throw e;
+        if (response.status === 404) {
+            return new RobotsTxtFile({
+                isAllowed() {
+                    return true;
+                },
+                getSitemaps() {
+                    return [];
+                },
+            }, proxyUrl, logger);
         }
+        // @ts-ignore
+        return new RobotsTxtFile(robotsParser(url.toString(), await response.text()), proxyUrl, logger);
     }
     /**
      * Check if a URL should be crawled by robots.
@@ -82,19 +81,19 @@ export class RobotsTxtFile {
      * @param [userAgent] relevant user agent, default to `*`
      */
     isAllowed(url, userAgent = '*') {
-        return this.robots.isAllowed(url, userAgent) ?? true; // `undefined` means that there is no explicit rule for the requested URL - assume it's allowed
+        return this.#robots.isAllowed(url, userAgent) ?? true; // `undefined` means that there is no explicit rule for the requested URL - assume it's allowed
     }
     /**
      * Get URLs of sitemaps referenced in the robots file.
      */
     getSitemaps() {
-        return this.robots.getSitemaps();
+        return this.#robots.getSitemaps();
     }
     /**
      * Parse all the sitemaps referenced in the robots file.
      */
     async parseSitemaps() {
-        return Sitemap.load(this.robots.getSitemaps(), this.proxyUrl);
+        return Sitemap.load(this.#robots.getSitemaps(), this.#proxyUrl, { logger: this.#logger });
     }
     /**
      * Get all URLs from all the sitemaps referenced in the robots file. A shorthand for `(await robots.parseSitemaps()).urls`.
@@ -103,6 +102,3 @@ export class RobotsTxtFile {
         return (await this.parseSitemaps()).urls;
     }
 }
-// to stay backwards compatible
-export { RobotsTxtFile as RobotsFile };
-//# sourceMappingURL=robots.js.map

package/internals/sitemap.d.ts

@@ -1,5 +1,4 @@
-// @ts-ignore optional peer dependency or compatibility with es2022
-import type { Delays } from 'got-scraping';
+import type { BaseHttpClient, CrawleeLogger } from '@crawlee/types';
 interface SitemapUrlData {
     loc: string;
     lastmod?: Date;
@@ -36,9 +35,29 @@ export interface ParseSitemapOptions {
      */
     sitemapRetries?: number;
     /**
-     * Network timeouts for sitemap fetching. See [Got documentation](https://github.com/sindresorhus/got/blob/main/documentation/6-timeout.md) for more details.
+     * Timeout settings for network requests when fetching sitemaps. By default this is `30000` milliseconds (30 seconds).
      */
-    networkTimeouts?: Delays;
+    timeoutMillis?: number;
+    /**
+     * If true, the parser will log a warning if it fails to fetch a sitemap due to a network error
+     * @default true
+     */
+    reportNetworkErrors?: boolean;
+    /**
+     * Custom HTTP client to be used for fetching sitemaps.
+     */
+    httpClient?: BaseHttpClient;
+    /**
+     * Optional filter for nested sitemap URLs discovered in sitemap index files.
+     * Called with the URL of each child sitemap before it is fetched.
+     * Return `true` to include the sitemap, `false` to skip it.
+     * If not provided, all nested sitemaps are followed.
+     */
+    nestedSitemapFilter?: (sitemapUrl: string) => boolean;
+    /**
+     * Optional logger for reporting warnings during sitemap parsing.
+     */
+    logger?: CrawleeLogger;
 }
 export declare function parseSitemap<T extends ParseSitemapOptions>(initialSources: SitemapSource[], proxyUrl?: string, options?: T): AsyncIterable<T['emitNestedSitemaps'] extends true ? SitemapUrl | NestedSitemap : SitemapUrl>;
 /**
@@ -62,7 +81,7 @@ export declare class Sitemap {
      * @param url The domain URL to fetch the sitemap for.
      * @param proxyUrl A proxy to be used for fetching the sitemap file.
      */
-    static tryCommonNames(url: string, proxyUrl?: string): Promise<Sitemap>;
+    static tryCommonNames(url: string, proxyUrl?: string, parseSitemapOptions?: ParseSitemapOptions): Promise<Sitemap>;
     /**
      * Fetch sitemap content from given URL or URLs and return URLs of referenced pages.
      * @param urls sitemap URL(s)
@@ -74,8 +93,46 @@ export declare class Sitemap {
      * @param content XML sitemap content
      * @param proxyUrl URL of a proxy to be used for fetching sitemap contents
      */
-    static fromXmlString(content: string, proxyUrl?: string): Promise<Sitemap>;
-    protected static parse(sources: SitemapSource[], proxyUrl?: string, parseSitemapOptions?: ParseSitemapOptions): Promise<Sitemap>;
+    static fromXmlString(content: string, proxyUrl?: string, parseSitemapOptions?: ParseSitemapOptions): Promise<Sitemap>;
+    private static parse;
 }
+/**
+ * Given a list of URLs, discover related sitemap files for these domains by checking the `robots.txt` file,
+ * the default `sitemap.xml` & `sitemap.txt` files and the URLs themselves.
+ * @param `urls` The list of URLs to discover sitemaps for.
+ * @param `options` Options for sitemap discovery
+ * @returns An async iterable with the discovered sitemap URLs.
+ */
+export declare function discoverValidSitemaps(urls: string[], options?: {
+    /**
+     * Proxy URL to be used for network requests.
+     */
+    proxyUrl?: string;
+    /**
+     * Timeout in milliseconds for the entire `discoverValidSitemaps` call.
+     * An `AbortController` is created internally and its signal is passed to every HTTP request,
+     * so the whole discovery operation is cancelled once the timeout elapses.
+     * Defaults to `60_000` ms (60 seconds) to prevent indefinite hangs.
+     */
+    timeoutMillis?: number;
+    /**
+     * An external `AbortSignal` to cancel the entire discovery operation.
+     * If both `signal` and `timeout` are provided, the operation is cancelled
+     * when either the signal is aborted or the timeout elapses (whichever comes first).
+     */
+    signal?: AbortSignal;
+    /**
+     * Timeout in milliseconds for each individual HTTP request during discovery.
+     * Defaults to `20000` ms (20 seconds).
+     */
+    requestTimeoutMillis?: number;
+    /**
+     * HTTP client to be used for network requests.
+     */
+    httpClient?: BaseHttpClient;
+    /**
+     * Optional logger for reporting warnings during sitemap discovery.
+     */
+    logger?: CrawleeLogger;
+}): AsyncIterable<string>;
 export {};
-//# sourceMappingURL=sitemap.d.ts.map

package/internals/sitemap.js

@@ -2,29 +2,31 @@ import { createHash } from 'node:crypto';
 import { PassThrough, pipeline, Readable, Transform } from 'node:stream';
 import { StringDecoder } from 'node:string_decoder';
 import { createGunzip } from 'node:zlib';
+import { FetchHttpClient } from '@crawlee/http-client';
 import sax from 'sax';
 import MIMEType from 'whatwg-mimetype';
-import log from '@apify/log';
+import { mergeAsyncIterables } from './iterables.js';
+import { RobotsTxtFile } from './robots.js';
 class SitemapTxtParser extends Transform {
-    decoder = new StringDecoder('utf8');
-    buffer = '';
+    #decoder = new StringDecoder('utf8');
+    #buffer = '';
     constructor() {
         super({
             readableObjectMode: true,
             transform: (chunk, _encoding, callback) => {
-                this.processBuffer(this.decoder.write(chunk), false);
+                this.processBuffer(this.#decoder.write(chunk), false);
                 callback();
             },
             flush: (callback) => {
-                this.processBuffer(this.decoder.end(), true);
+                this.processBuffer(this.#decoder.end(), true);
                 callback();
             },
         });
     }
     processBuffer(input, finalize) {
-        this.buffer += input;
-        if (finalize || this.buffer.includes('\n')) {
-            const parts = this.buffer
+        this.#buffer += input;
+        if (finalize || this.#buffer.includes('\n')) {
+            const parts = this.#buffer
                 .split('\n')
                 .map((part) => part.trim())
                 .filter((part) => part.length > 0);
@@ -32,98 +34,101 @@ class SitemapTxtParser extends Transform {
                 for (const url of parts) {
                     this.push({ type: 'url', loc: url });
                 }
-                this.buffer = '';
+                this.#buffer = '';
             }
             else if (parts.length > 0) {
                 for (const url of parts.slice(0, -1)) {
                     this.push({ type: 'url', loc: url });
                 }
-                this.buffer = parts.at(-1);
+                this.#buffer = parts.at(-1);
             }
         }
     }
 }
 class SitemapXmlParser extends Transform {
-    decoder = new StringDecoder('utf8');
-    parser = new sax.SAXParser(true);
-    rootTagName;
-    currentTag = undefined;
-    url = {};
+    #decoder = new StringDecoder('utf8');
+    #parser = new sax.SAXParser(true);
+    #rootTagName;
+    #currentTag = undefined;
+    #url = {};
     constructor() {
         super({
             readableObjectMode: true,
             transform: (chunk, _encoding, callback) => {
-                this.parser.write(this.decoder.write(chunk));
+                this.#parser.write(this.#decoder.write(chunk));
                 callback();
             },
             flush: (callback) => {
-                const rest = this.decoder.end();
+                const rest = this.#decoder.end();
                 if (rest.length > 0) {
-                    this.parser.write(rest);
+                    this.#parser.write(rest);
                 }
-                this.parser.end();
+                this.#parser.end();
                 callback();
             },
         });
-        this.parser.onopentag = this.onOpenTag.bind(this);
-        this.parser.onclosetag = this.onCloseTag.bind(this);
-        this.parser.ontext = this.onText.bind(this);
-        this.parser.oncdata = this.onText.bind(this);
-        this.parser.onerror = this.destroy.bind(this);
+        this.#parser.onopentag = this.onOpenTag.bind(this);
+        this.#parser.onclosetag = this.onCloseTag.bind(this);
+        this.#parser.ontext = this.onText.bind(this);
+        this.#parser.oncdata = this.onText.bind(this);
+        this.#parser.onerror = this.destroy.bind(this);
     }
     onOpenTag(node) {
-        if (this.rootTagName !== undefined) {
+        if (this.#rootTagName !== undefined) {
             if (node.name === 'loc' ||
                 node.name === 'lastmod' ||
                 node.name === 'priority' ||
                 node.name === 'changefreq') {
-                this.currentTag = node.name;
+                this.#currentTag = node.name;
             }
         }
         if (node.name === 'urlset') {
-            this.rootTagName = 'urlset';
+            this.#rootTagName = 'urlset';
         }
         if (node.name === 'sitemapindex') {
-            this.rootTagName = 'sitemapindex';
+            this.#rootTagName = 'sitemapindex';
         }
     }
     onCloseTag(name) {
         if (name === 'loc' || name === 'lastmod' || name === 'priority' || name === 'changefreq') {
-            this.currentTag = undefined;
+            this.#currentTag = undefined;
         }
-        if (name === 'url' && this.url.loc !== undefined) {
-            this.push({ type: 'url', ...this.url, loc: this.url.loc });
-            this.url = {};
+        if (name === 'url') {
+            if (this.#url.loc !== undefined) {
+                this.push({ type: 'url', ...this.#url, loc: this.#url.loc });
+            }
+            this.#url = {};
         }
     }
     onText(text) {
-        if (this.currentTag === 'loc') {
-            if (this.rootTagName === 'sitemapindex') {
+        if (this.#currentTag === 'loc') {
+            if (this.#rootTagName === 'sitemapindex') {
                 this.push({ type: 'sitemapUrl', url: text.trim() });
             }
-            if (this.rootTagName === 'urlset') {
-                this.url ??= {};
-                this.url.loc = text.trim();
+            if (this.#rootTagName === 'urlset') {
+                this.#url ??= {};
+                this.#url.loc = text.trim();
             }
         }
         text = text.trim();
-        if (this.currentTag === 'lastmod') {
-            this.url.lastmod = new Date(text);
+        if (this.#currentTag === 'lastmod') {
+            const lastmod = new Date(text);
+            if (!Number.isNaN(lastmod.getTime())) {
+                this.#url.lastmod = lastmod;
+            }
         }
-        if (this.currentTag === 'priority') {
-            this.url.priority = Number(text);
+        if (this.#currentTag === 'priority') {
+            this.#url.priority = Number(text);
         }
-        if (this.currentTag === 'changefreq') {
+        if (this.#currentTag === 'changefreq') {
             if (['always', 'hourly', 'daily', 'weekly', 'monthly', 'yearly', 'never'].includes(text)) {
-                this.url.changefreq = text;
+                this.#url.changefreq = text;
             }
         }
     }
 }
 export async function* parseSitemap(initialSources, proxyUrl, options) {
-    const { gotScraping } = await import('got-scraping');
-    const { fileTypeStream } = await import('file-type');
-    const { emitNestedSitemaps = false, maxDepth = Infinity, sitemapRetries = 3, networkTimeouts } = options ?? {};
+    const { httpClient = new FetchHttpClient(), emitNestedSitemaps = false, maxDepth = Infinity, sitemapRetries = 3, timeoutMillis: timeout = 30000, reportNetworkErrors = true, nestedSitemapFilter, logger, } = options ?? {};
     const sources = [...initialSources];
     const visitedSitemapUrls = new Set();
     const createParser = (contentType = '', url) => {
@@ -145,7 +150,6 @@ export async function* parseSitemap(initialSources, proxyUrl, options) {
     while (sources.length > 0) {
         const source = sources.shift();
         if ((source?.depth ?? 0) > maxDepth) {
-            log.debug(`Skipping sitemap ${source.type === 'url' ? source.url : ''} because it reached max depth ${maxDepth}.`);
             continue;
         }
         let items = null;
@@ -155,23 +159,29 @@ export async function* parseSitemap(initialSources, proxyUrl, options) {
             let retriesLeft = sitemapRetries + 1;
             while (retriesLeft-- > 0) {
                 try {
-                    const sitemapStream = await new Promise((resolve, reject) => {
-                        const request = gotScraping.stream({
-                            url: sitemapUrl,
-                            proxyUrl,
+                    let sitemapResponse;
+                    try {
+                        sitemapResponse = await httpClient.sendRequest(new Request(sitemapUrl, {
                             method: 'GET',
-                            timeout: networkTimeouts,
                             headers: {
-                                accept: 'text/plain, application/xhtml+xml, application/xml;q=0.9, */*;q=0.8',
+                                accept: '*/*',
                             },
+                        }), {
+                            proxyUrl,
+                            timeoutMillis: timeout,
                         });
-                        request.on('response', () => resolve(request));
-                        request.on('error', reject);
-                    });
+                    }
+                    catch (error) {
+                        sitemapResponse = null;
+                    }
                     let error = null;
-                    if (sitemapStream.response.statusCode >= 200 && sitemapStream.response.statusCode < 300) {
-                        let contentType = sitemapStream.response.headers['content-type'];
-                        const streamWithType = await fileTypeStream(sitemapStream);
+                    if (sitemapResponse && sitemapResponse.status >= 200 && sitemapResponse.status < 300) {
+                        let contentType = sitemapResponse.headers.get('content-type');
+                        if (sitemapResponse.body === null) {
+                            break;
+                        }
+                        const { fileTypeStream } = await import('file-type');
+                        const streamWithType = await fileTypeStream(Readable.fromWeb(sitemapResponse.body));
                         if (streamWithType.fileType !== undefined) {
                             contentType = streamWithType.fileType.mime;
                         }
@@ -184,29 +194,37 @@ export async function* parseSitemap(initialSources, proxyUrl, options) {
                                 sitemapUrl.pathname = sitemapUrl.pathname.substring(0, sitemapUrl.pathname.length - 3);
                             }
                         }
-                        items = pipeline(streamWithType, isGzipped ? createGunzip() : new PassThrough(), createParser(contentType, sitemapUrl), (e) => {
-                            if (e !== undefined) {
-                                error = e;
+                        items = pipeline(streamWithType, isGzipped ? createGunzip() : new PassThrough(), createParser(contentType ?? undefined, sitemapUrl), (e) => {
+                            if (e !== undefined && e !== null) {
+                                error = { type: 'parser', error: e };
                             }
                         });
                     }
                     else {
-                        error = new Error(`Failed to fetch sitemap: ${sitemapUrl}, status code: ${sitemapStream.response.statusCode}`);
+                        error = {
+                            type: 'fetch',
+                            error: new Error(`Failed to fetch sitemap: ${sitemapUrl}, status code: ${sitemapResponse?.status}`),
+                        };
                     }
                     if (error !== null) {
-                        throw error;
+                        const shouldIgnoreError = error.type === 'fetch' && !reportNetworkErrors;
+                        if (!shouldIgnoreError) {
+                            throw error.error;
+                        }
+                    }
+                    else {
+                        break;
                     }
-                    break;
                 }
                 catch (e) {
-                    log.warning(`Malformed sitemap content: ${sitemapUrl}, ${retriesLeft === 0 ? 'no retries left.' : 'retrying...'} (${e})`);
+                    logger?.warning(`Malformed sitemap content: ${sitemapUrl}, ${retriesLeft === 0 ? 'no retries left.' : 'retrying...'} (${e})`);
                 }
             }
         }
         else if (source.type === 'raw') {
             items = pipeline(Readable.from([source.content]), createParser('text/xml'), (error) => {
                 if (error !== undefined) {
-                    log.warning(`Malformed sitemap content: ${error}`);
+                    logger?.warning(`Malformed sitemap content: ${error}`);
                 }
             });
         }
@@ -215,6 +233,10 @@ export async function* parseSitemap(initialSources, proxyUrl, options) {
         }
         for await (const item of items) {
             if (item.type === 'sitemapUrl' && !visitedSitemapUrls.has(item.url)) {
+                if (nestedSitemapFilter && !nestedSitemapFilter(item.url)) {
+                    logger?.debug(`Skipping sitemap ${item.url} due to nestedSitemapFilter.`);
+                    continue;
+                }
                 sources.push({ type: 'url', url: item.url, depth: (source.depth ?? 0) + 1 });
                 if (emitNestedSitemaps) {
                     yield { loc: item.url, originSitemapUrl: null };
@@ -254,7 +276,7 @@ export class Sitemap {
      * @param url The domain URL to fetch the sitemap for.
      * @param proxyUrl A proxy to be used for fetching the sitemap file.
      */
-    static async tryCommonNames(url, proxyUrl) {
+    static async tryCommonNames(url, proxyUrl, parseSitemapOptions) {
         const sitemapUrls = [];
         const sitemapUrl = new URL(url);
         sitemapUrl.search = '';
@@ -262,7 +284,7 @@ export class Sitemap {
         sitemapUrls.push(sitemapUrl.toString());
         sitemapUrl.pathname = '/sitemap.txt';
         sitemapUrls.push(sitemapUrl.toString());
-        return Sitemap.load(sitemapUrls, proxyUrl);
+        return Sitemap.load(sitemapUrls, proxyUrl, { reportNetworkErrors: false, ...parseSitemapOptions });
     }
     /**
      * Fetch sitemap content from given URL or URLs and return URLs of referenced pages.
@@ -277,8 +299,8 @@ export class Sitemap {
      * @param content XML sitemap content
      * @param proxyUrl URL of a proxy to be used for fetching sitemap contents
      */
-    static async fromXmlString(content, proxyUrl) {
-        return await this.parse([{ type: 'raw', content }], proxyUrl);
+    static async fromXmlString(content, proxyUrl, parseSitemapOptions) {
+        return await this.parse([{ type: 'raw', content }], proxyUrl, parseSitemapOptions);
     }
     static async parse(sources, proxyUrl, parseSitemapOptions) {
         const urls = [];
@@ -287,10 +309,128 @@ export class Sitemap {
                 urls.push(item.loc);
             }
         }
-        catch {
+        catch (e) {
+            parseSitemapOptions?.logger?.warning(`Sitemap.load: Failed to load sitemap, returning empty result. (${e})`);
             return new Sitemap([]);
         }
         return new Sitemap(urls);
     }
 }
-//# sourceMappingURL=sitemap.js.map
+/**
+ * Given a list of URLs, discover related sitemap files for these domains by checking the `robots.txt` file,
+ * the default `sitemap.xml` & `sitemap.txt` files and the URLs themselves.
+ * @param `urls` The list of URLs to discover sitemaps for.
+ * @param `options` Options for sitemap discovery
+ * @returns An async iterable with the discovered sitemap URLs.
+ */
+export async function* discoverValidSitemaps(urls, options = {}) {
+    const { proxyUrl, timeoutMillis = 60_000, signal: externalSignal, requestTimeoutMillis = 20_000, httpClient = new FetchHttpClient(), logger, } = options;
+    const controller = new AbortController();
+    const timeoutHandle = setTimeout(() => controller.abort(), timeoutMillis);
+    const onExternalAbort = () => controller.abort();
+    if (externalSignal) {
+        if (externalSignal.aborted) {
+            controller.abort();
+        }
+        else {
+            externalSignal.addEventListener('abort', onExternalAbort, { once: true });
+        }
+    }
+    const signal = controller.signal;
+    const sitemapUrls = new Set();
+    const addSitemapUrl = (url) => {
+        const sizeBefore = sitemapUrls.size;
+        sitemapUrls.add(url);
+        if (sitemapUrls.size > sizeBefore) {
+            return url;
+        }
+        return undefined;
+    };
+    const urlExists = async (url) => {
+        if (!httpClient) {
+            return false;
+        }
+        try {
+            const response = await httpClient.sendRequest(new Request(url, { method: 'HEAD' }), {
+                proxyUrl,
+                timeoutMillis: requestTimeoutMillis,
+                signal,
+            });
+            return response.status >= 200 && response.status < 400;
+        }
+        catch {
+            return false;
+        }
+    };
+    const discoverSitemapsForDomainUrls = async function* (hostname, domainUrls) {
+        if (!hostname) {
+            return;
+        }
+        try {
+            const robotsFile = await RobotsTxtFile.find(domainUrls[0], {
+                proxyUrl,
+                timeoutMillis: requestTimeoutMillis,
+                signal,
+                httpClient,
+                logger,
+            });
+            for (const sitemapUrl of robotsFile.getSitemaps()) {
+                if (addSitemapUrl(sitemapUrl)) {
+                    yield sitemapUrl;
+                }
+            }
+        }
+        catch (err) {
+            logger?.warning(`Failed to fetch robots.txt file for ${hostname}`, { error: err });
+        }
+        const sitemapUrl = domainUrls.find((url) => /sitemap\.(?:xml|txt)(?:\.gz)?$/i.test(url));
+        if (sitemapUrl !== undefined) {
+            if (addSitemapUrl(sitemapUrl)) {
+                yield sitemapUrl;
+            }
+        }
+        else {
+            const firstUrl = new URL(domainUrls[0]);
+            const possibleSitemapPathnames = ['/sitemap.xml', '/sitemap.txt', '/sitemap_index.xml'];
+            const candidateSitemapUrls = possibleSitemapPathnames.map((pathname) => {
+                firstUrl.pathname = pathname;
+                return firstUrl.toString();
+            });
+            const candidateResults = await Promise.allSettled(candidateSitemapUrls.map(urlExists));
+            for (const [index, result] of candidateResults.entries()) {
+                const candidateSitemapUrl = candidateSitemapUrls[index];
+                if (result.status === 'fulfilled') {
+                    if (result.value && addSitemapUrl(candidateSitemapUrl)) {
+                        yield candidateSitemapUrl;
+                    }
+                }
+                else {
+                    logger?.debug(`Failed to check sitemap candidate ${candidateSitemapUrl} for ${hostname}`, {
+                        error: result.reason,
+                    });
+                }
+            }
+        }
+    };
+    const groupedUrls = urls.reduce((acc, url) => {
+        const hostname = new URL(url)?.hostname ?? '';
+        acc[hostname] ??= [];
+        acc[hostname].push(url);
+        return acc;
+    }, {});
+    const iterables = Object.entries(groupedUrls).map(([hostname, domainUrls]) => discoverSitemapsForDomainUrls(hostname, domainUrls));
+    const discoveredUrls = new Set();
+    try {
+        for await (const url of mergeAsyncIterables(...iterables)) {
+            if (discoveredUrls.has(url)) {
+                continue;
+            }
+            discoveredUrls.add(url);
+            yield url;
+        }
+    }
+    finally {
+        clearTimeout(timeoutHandle);
+        externalSignal?.removeEventListener('abort', onExternalAbort);
+    }
+}