@crawlee/playwright 4.0.0-beta.2 → 4.0.0-beta.200

package/internals/enqueue-links/click-elements.d.ts

@@ -1,4 +1,4 @@
-import type { GlobInput, PseudoUrlInput, RegExpInput, RequestProvider, RequestTransform } from '@crawlee/browser';
+import type { IRequestManager, RequestTransform, SkippedRequestCallback, UrlPatternInput } from '@crawlee/browser';
 import type { BatchAddRequestsResult, Dictionary } from '@crawlee/types';
 // @ts-ignore optional peer dependency or compatibility with es2022
 import type { Page } from 'playwright';
@@ -9,9 +9,9 @@ export interface EnqueueLinksByClickingElementsOptions {
      */
     page: Page;
     /**
-     * A request queue to which the URLs will be enqueued.
+     * * A request manager to which the URLs will be enqueued.
      */
-    requestQueue: RequestProvider;
+    requestManager: IRequestManager;
     /**
      * A CSS selector matching elements to be clicked on. Unlike in {@link enqueueLinks}, there is no default
      * value. This is to prevent suboptimal use of this function by using it too broadly.
@@ -26,72 +26,50 @@ export interface EnqueueLinksByClickingElementsOptions {
      */
     clickOptions?: ClickOptions;
     /**
-     * An array of glob pattern strings or plain objects
-     * containing glob pattern strings matching the URLs to be enqueued.
+     * An array of URL patterns that URLs must match to be enqueued.
      *
-     * The plain objects must include at least the `glob` property, which holds the glob pattern string.
-     * All remaining keys will be used as request options for the corresponding enqueued {@link Request} objects.
+     * Accepts glob pattern strings, `{ glob: string }` objects, `RegExp` instances, or `{ regexp: RegExp }` objects.
      *
-     * The matching is always case-insensitive.
-     * If you need case-sensitive matching, use `regexps` property directly.
+     * Glob matching is always case-insensitive.
+     * If you need case-sensitive matching, use a `RegExp`.
      *
-     * If `globs` is an empty array or `undefined`, then the function
+     * If `include` is an empty array or `undefined`, then the function
      * enqueues all the intercepted navigation requests produced by the page
      * after clicking on elements matching the provided CSS selector.
      */
-    globs?: GlobInput[];
+    include?: UrlPatternInput[];
     /**
-     * An array of regular expressions or plain objects
-     * containing regular expressions matching the URLs to be enqueued.
+     * An array of URL patterns. Matching URLs will **not** be enqueued.
      *
-     * The plain objects must include at least the `regexp` property, which holds the regular expression.
-     * All remaining keys will be used as request options for the corresponding enqueued {@link Request} objects.
+     * Accepts glob pattern strings, `{ glob: string }` objects, `RegExp` instances, or `{ regexp: RegExp }` objects.
      *
-     * If `regexps` is an empty array or `undefined`, then the function
-     * enqueues all the intercepted navigation requests produced by the page
-     * after clicking on elements matching the provided CSS selector.
-     */
-    regexps?: RegExpInput[];
-    /**
-     * *NOTE:* In future versions of SDK the options will be removed.
-     * Please use `globs` or `regexps` instead.
-     *
-     * An array of {@link PseudoUrl} strings or plain objects
-     * containing {@link PseudoUrl} strings matching the URLs to be enqueued.
-     *
-     * The plain objects must include at least the `purl` property, which holds the pseudo-URL pattern string.
-     * All remaining keys will be used as request options for the corresponding enqueued {@link Request} objects.
-     *
-     * With a pseudo-URL string, the matching is always case-insensitive.
-     * If you need case-sensitive matching, use `regexps` property directly.
-     *
-     * If `pseudoUrls` is an empty array or `undefined`, then the function
-     * enqueues all the intercepted navigation requests produced by the page
-     * after clicking on elements matching the provided CSS selector.
-     *
-     * @deprecated prefer using `globs` or `regexps` instead
+     * Glob matching is always case-insensitive.
+     * If you need case-sensitive matching, use a `RegExp`.
      */
-    pseudoUrls?: PseudoUrlInput[];
+    exclude?: readonly UrlPatternInput[];
     /**
-     * Just before a new {@link Request} is constructed and enqueued to the {@link RequestQueue}, this function can be used
-     * to remove it or modify its contents such as `userData`, `payload` or, most importantly `uniqueKey`. This is useful
-     * when you need to enqueue multiple `Requests` to the queue that share the same URL, but differ in methods or payloads,
-     * or to dynamically update or create `userData`.
-     *
-     * For example: by adding `useExtendedUniqueKey: true` to the `request` object, `uniqueKey` will be computed from
-     * a combination of `url`, `method` and `payload` which enables crawling of websites that navigate using form submits
-     * (POST requests).
+     * After request options are filtered by `include`/`exclude` patterns,
+     * this function can be used to remove them or modify their contents such as `userData`, `payload` or, most importantly
+     * `uniqueKey`. This is useful when you need to enqueue multiple `Requests` to the queue that share the same URL,
+     * but differ in methods or payloads, or to dynamically update or create `userData`.
      *
      * **Example:**
      * ```javascript
      * {
      *     transformRequestFunction: (request) => {
      *         request.userData.foo = 'bar';
-     *         request.useExtendedUniqueKey = true;
      *         return request;
      *     }
      * }
      * ```
+     *
+     * Note that `transformRequestFunction` has the highest priority and can overwrite
+     * the global `label` option.
+     *
+     * The function receives a {@link RequestOptions} object and can return either:
+     * - The modified {@link RequestOptions} object
+     * - `'unchanged'` to keep the original options as-is
+     * - A falsy value or `'skip'` to exclude the request from the queue
      */
     transformRequestFunction?: RequestTransform;
     /**
@@ -131,6 +109,12 @@ export interface EnqueueLinksByClickingElementsOptions {
      * @default false
      */
     skipNavigation?: boolean;
+    /**
+     * When a request is skipped for some reason, you can use this callback to act on it.
+     * This is fired for requests skipped because they don't match enqueueLinks filters
+     * or because they were removed by `transformRequestFunction`.
+     */
+    onSkippedRequest?: SkippedRequestCallback;
 }
 /**
  * The function finds elements matching a specific CSS selector in a Playwright page,
@@ -141,8 +125,7 @@ export interface EnqueueLinksByClickingElementsOptions {
  * in `href` elements, but rather navigations are triggered in click handlers.
  * If you're looking to find URLs in `href` attributes of the page, see {@link enqueueLinks}.
  *
- * Optionally, the function allows you to filter the target links' URLs using an array of {@link PseudoUrl} objects
- * and override settings of the enqueued {@link Request} objects.
+ * Optionally, the function allows you to filter the target links' URLs using an array of glob or regexp patterns.
  *
  * **IMPORTANT**: To be able to do this, this function uses various mutations on the page,
  * such as changing the Z-index of elements being clicked and their visibility. Therefore,
@@ -162,11 +145,11 @@ export interface EnqueueLinksByClickingElementsOptions {
  * ```javascript
  * await playwrightUtils.enqueueLinksByClickingElements({
  *   page,
- *   requestQueue,
+ *   requestManager,
  *   selector: 'a.product-detail',
- *   pseudoUrls: [
- *       'https://www.example.com/handbags/[.*]'
- *       'https://www.example.com/purses/[.*]'
+ *   include: [
+ *       'https://www.example.com/handbags/*',
+ *       'https://www.example.com/purses/*',
  *   ],
  * });
  * ```
@@ -200,4 +183,3 @@ export declare function clickElementsAndInterceptNavigationRequests(options: Cli
  */
 export declare function clickElements(page: Page, selector: string, clickOptions?: ClickOptions): Promise<void>;
 export {};
-//# sourceMappingURL=click-elements.d.ts.map

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

@@ -1,9 +1,25 @@
 import { URL } from 'node:url';
-import { constructGlobObjectsFromGlobs, constructRegExpObjectsFromPseudoUrls, constructRegExpObjectsFromRegExps, createRequestOptions, createRequests, } from '@crawlee/browser';
-import ow from 'ow';
-import log_ from '@apify/log';
+import { applyRequestTransform, constructUrlPatternObjects, createRequestOptions, createSkippedRequestArgs, 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 log = log_.child({ prefix: 'Playwright Click Elements' });
+const getLog = () => serviceLocator.getChildLog('Playwright Click Elements');
+const enqueueLinksByClickingElementsOptionsSchema = z.strictObject({
+    page: schemas.objectWithKeys(['goto', 'evaluate']),
+    requestManager: schemas.objectWithKeys(['fetchNextRequest', 'addRequestsBatched']),
+    selector: z.string(),
+    userData: schemas.anyObject.optional(),
+    clickOptions: schemas.anyObject.optional(),
+    include: schemas.arrayOf(urlPatternSchema, 'URL patterns').optional(),
+    exclude: schemas.arrayOf(urlPatternSchema, 'URL patterns').optional(),
+    transformRequestFunction: schemas.anyFunction.optional(),
+    waitForPageIdleSecs: schemas.anyNumber.default(1),
+    maxWaitForPageIdleSecs: schemas.anyNumber.default(5),
+    label: z.string().optional(),
+    forefront: z.boolean().optional(),
+    skipNavigation: z.boolean().optional(),
+    onSkippedRequest: schemas.anyFunction.optional(),
+});
 /**
  * The function finds elements matching a specific CSS selector in a Playwright page,
  * clicks all those elements using a mouse move and a left mouse button click and intercepts
@@ -13,8 +29,7 @@ const log = log_.child({ prefix: 'Playwright Click Elements' });
  * in `href` elements, but rather navigations are triggered in click handlers.
  * If you're looking to find URLs in `href` attributes of the page, see {@link enqueueLinks}.
  *
- * Optionally, the function allows you to filter the target links' URLs using an array of {@link PseudoUrl} objects
- * and override settings of the enqueued {@link Request} objects.
+ * Optionally, the function allows you to filter the target links' URLs using an array of glob or regexp patterns.
  *
  * **IMPORTANT**: To be able to do this, this function uses various mutations on the page,
  * such as changing the Z-index of elements being clicked and their visibility. Therefore,
@@ -34,11 +49,11 @@ const log = log_.child({ prefix: 'Playwright Click Elements' });
  * ```javascript
  * await playwrightUtils.enqueueLinksByClickingElements({
  *   page,
- *   requestQueue,
+ *   requestManager,
  *   selector: 'a.product-detail',
- *   pseudoUrls: [
- *       'https://www.example.com/handbags/[.*]'
- *       'https://www.example.com/purses/[.*]'
+ *   include: [
+ *       'https://www.example.com/handbags/*',
+ *       'https://www.example.com/purses/*',
  *   ],
  * });
  * ```
@@ -46,36 +61,13 @@ const log = log_.child({ prefix: 'Playwright Click Elements' });
  * @returns Promise that resolves to {@link BatchAddRequestsResult} object.
  */
 export async function enqueueLinksByClickingElements(options) {
-    ow(options, ow.object.exactShape({
-        page: ow.object.hasKeys('goto', 'evaluate'),
-        requestQueue: ow.object.hasKeys('fetchNextRequest', 'addRequest'),
-        selector: ow.string,
-        userData: ow.optional.object,
-        clickOptions: ow.optional.object.hasKeys('clickCount', 'delay'),
-        pseudoUrls: ow.optional.array.ofType(ow.any(ow.string, ow.object.hasKeys('purl'))),
-        globs: ow.optional.array.ofType(ow.any(ow.string, ow.object.hasKeys('glob'))),
-        regexps: ow.optional.array.ofType(ow.any(ow.regExp, ow.object.hasKeys('regexp'))),
-        transformRequestFunction: ow.optional.function,
-        waitForPageIdleSecs: ow.optional.number,
-        maxWaitForPageIdleSecs: ow.optional.number,
-        label: ow.optional.string,
-        forefront: ow.optional.boolean,
-        skipNavigation: ow.optional.boolean,
-    }));
-    const { page, requestQueue, selector, clickOptions, pseudoUrls, globs, regexps, transformRequestFunction, waitForPageIdleSecs = 1, maxWaitForPageIdleSecs = 5, forefront, } = options;
+    const parsedOptions = parseArgument(options, enqueueLinksByClickingElementsOptionsSchema, 'EnqueueLinksByClickingElementsOptions');
+    const { page, requestManager, selector, clickOptions, include, exclude, transformRequestFunction, waitForPageIdleSecs, maxWaitForPageIdleSecs, forefront, onSkippedRequest, } = parsedOptions;
     const waitForPageIdleMillis = waitForPageIdleSecs * 1000;
     const maxWaitForPageIdleMillis = maxWaitForPageIdleSecs * 1000;
-    const urlPatternObjects = [];
-    if (pseudoUrls?.length) {
-        log.deprecated('`pseudoUrls` option is deprecated, use `globs` or `regexps` instead');
-        urlPatternObjects.push(...constructRegExpObjectsFromPseudoUrls(pseudoUrls));
-    }
-    if (globs?.length) {
-        urlPatternObjects.push(...constructGlobObjectsFromGlobs(globs));
-    }
-    if (regexps?.length) {
-        urlPatternObjects.push(...constructRegExpObjectsFromRegExps(regexps));
-    }
+    const hasOnSkippedRequest = onSkippedRequest !== undefined;
+    const urlExcludePatternObjects = exclude?.length ? constructUrlPatternObjects(exclude) : [];
+    const urlPatternObjects = include?.length ? constructUrlPatternObjects(include) : [];
     const interceptedRequests = await clickElementsAndInterceptNavigationRequests({
         page,
         selector,
@@ -83,12 +75,21 @@ export async function enqueueLinksByClickingElements(options) {
         maxWaitForPageIdleMillis,
         clickOptions,
     });
-    let requestOptions = createRequestOptions(interceptedRequests, options);
+    const requestOptions = createRequestOptions(interceptedRequests, parsedOptions);
+    const skippedByFilters = [];
+    let filteredOptions = filterRequestOptionsByPatterns(requestOptions, urlPatternObjects.length > 0 ? urlPatternObjects : undefined, urlExcludePatternObjects, undefined, hasOnSkippedRequest ? (opts) => skippedByFilters.push(opts) : undefined);
+    if (onSkippedRequest && skippedByFilters.length > 0) {
+        await Promise.all(skippedByFilters.map(async (opts) => onSkippedRequest(createSkippedRequestArgs(opts, 'filters'))));
+    }
     if (transformRequestFunction) {
-        requestOptions = requestOptions.map(transformRequestFunction).filter((r) => !!r);
+        const skippedByTransform = [];
+        filteredOptions = applyRequestTransform(filteredOptions, transformRequestFunction, hasOnSkippedRequest ? (r) => skippedByTransform.push(r) : undefined);
+        if (onSkippedRequest && skippedByTransform.length > 0) {
+            await Promise.all(skippedByTransform.map(async (r) => onSkippedRequest(createSkippedRequestArgs(r, 'transform'))));
+        }
     }
-    const requests = createRequests(requestOptions, urlPatternObjects);
-    const { addedRequests } = await requestQueue.addRequestsBatched(requests, { forefront });
+    const requests = filteredOptions.map((opts) => new CrawleeRequest(opts));
+    const { addedRequests } = await requestManager.addRequestsBatched(requests, { forefront });
     return { processedRequests: addedRequests, unprocessedRequests: [] };
 }
 /**
@@ -114,7 +115,7 @@ export async function clickElementsAndInterceptNavigationRequests(options) {
     await restoreHistoryNavigationAndSaveCapturedUrls(page, uniqueRequests);
     // browser.off(BrowserEmittedEvents.TargetCreated, onTargetCreated);
     page.off('framenavigated', onFrameNavigated);
-    await context.unroute('*', onInterceptedRequest);
+    await context.unroute('**', onInterceptedRequest);
     const serializedRequests = Array.from(uniqueRequests);
     return serializedRequests.map((r) => JSON.parse(r));
 }
@@ -150,7 +151,9 @@ function createTargetCreatedHandler(requests) {
             await popup.close();
         }
         catch (err) {
-            log.debug('enqueueLinksByClickingElements: Could not close spawned page.', { error: err.stack });
+            getLog().debug('enqueueLinksByClickingElements: Could not close spawned page.', {
+                error: err.stack,
+            });
         }
     };
 }
@@ -158,7 +161,16 @@ function createTargetCreatedHandler(requests) {
  * @ignore
  */
 function isTopFrameNavigationRequest(page, req) {
-    return req.isNavigationRequest() && req.frame() === page.mainFrame();
+    try {
+        return req.isNavigationRequest() && req.frame() === page.mainFrame();
+    }
+    catch {
+        // `req.frame()` throws when the owning frame is unavailable - e.g. the request was
+        // issued by a service worker, or before/after its frame existed (see #3216). Such a
+        // request is not a top-frame navigation, so swallow the throw and let it pass through
+        // instead of crashing the route handler (which would leave the route unhandled).
+        return false;
+    }
 }
 /**
  * @ignore
@@ -223,7 +235,7 @@ function updateElementCssToEnableMouseClick(el, zIndex) {
  */
 export async function clickElements(page, selector, clickOptions) {
     const elementHandles = await page.$$(selector);
-    log.debug(`enqueueLinksByClickingElements: There are ${elementHandles.length} elements to click.`);
+    getLog().debug(`enqueueLinksByClickingElements: There are ${elementHandles.length} elements to click.`);
     let clickedElementsCount = 0;
     let zIndex = STARTING_Z_INDEX;
     let shouldLogWarning = true;
@@ -236,15 +248,15 @@ export async function clickElements(page, selector, clickOptions) {
         catch (err) {
             const e = err;
             if (shouldLogWarning && e.stack.includes('is detached from document')) {
-                log.warning(`An element with selector ${selector} that you're trying to click has been removed from the page. ` +
+                getLog().warning(`An element with selector ${selector} that you're trying to click has been removed from the page. ` +
                     'This was probably caused by an earlier click which triggered some JavaScript on the page that caused it to change. ' +
                     'If you\'re trying to enqueue pagination links, we suggest using the "next" button, if available and going one by one.');
                 shouldLogWarning = false;
             }
-            log.debug('enqueueLinksByClickingElements: Click failed.', { stack: e.stack });
+            getLog().debug('enqueueLinksByClickingElements: Click failed.', { stack: e.stack });
         }
     }
-    log.debug(`enqueueLinksByClickingElements: Successfully clicked ${clickedElementsCount} elements out of ${elementHandles.length}`);
+    getLog().debug(`enqueueLinksByClickingElements: Successfully clicked ${clickedElementsCount} elements out of ${elementHandles.length}`);
 }
 /**
  * This function tracks whether any requests, frame navigations or targets were emitted
@@ -263,8 +275,6 @@ export async function clickElements(page, selector, clickOptions) {
 async function waitForPageIdle({ page, waitForPageIdleMillis, maxWaitForPageIdleMillis, }) {
     return new Promise((resolve) => {
         let timeout;
-        let maxTimeout;
-        page.on('popup', activityHandler);
         function activityHandler() {
             clearTimeout(timeout);
             timeout = setTimeout(() => {
@@ -273,7 +283,7 @@ async function waitForPageIdle({ page, waitForPageIdleMillis, maxWaitForPageIdle
             }, waitForPageIdleMillis);
         }
         function maxTimeoutHandler() {
-            log.debug(`enqueueLinksByClickingElements: Page still showed activity after ${maxWaitForPageIdleMillis}ms. ` +
+            getLog().debug(`enqueueLinksByClickingElements: Page still showed activity after ${maxWaitForPageIdleMillis}ms. ` +
                 'This is probably due to the website itself dispatching requests, but some links may also have been missed.');
             finish();
         }
@@ -281,7 +291,8 @@ async function waitForPageIdle({ page, waitForPageIdleMillis, maxWaitForPageIdle
             page.off('request', activityHandler).off('framenavigated', activityHandler).off('popup', activityHandler);
             resolve();
         }
-        maxTimeout = setTimeout(maxTimeoutHandler, maxWaitForPageIdleMillis);
+        const maxTimeout = setTimeout(maxTimeoutHandler, maxWaitForPageIdleMillis);
+        page.on('popup', activityHandler);
         activityHandler(); // We call this once manually in case there would be no requests at all.
         page.on('request', activityHandler);
         page.on('framenavigated', activityHandler);
@@ -304,8 +315,7 @@ async function restoreHistoryNavigationAndSaveCapturedUrls(page, requests) {
             requests.add(JSON.stringify({ url }));
         }
         catch (err) {
-            log.debug('enqueueLinksByClickingElements: Failed to ', { error: err.stack });
+            getLog().debug('enqueueLinksByClickingElements: Failed to ', { error: err.stack });
         }
     });
 }
-//# sourceMappingURL=click-elements.js.map

package/internals/playwright-browser-pool.d.ts

@@ -0,0 +1,71 @@
+import type { Configuration } from '@crawlee/browser';
+import type { BrowserPool, BrowserPoolHooks, BrowserPoolOptions, PlaywrightPlugin, RemoteBrowserPool, RemoteBrowserPoolOptions } from '@crawlee/browser-pool';
+// @ts-ignore optional peer dependency or compatibility with es2022
+import type { Page } from 'playwright';
+import type { PlaywrightLaunchContext } from './playwright-launcher.js';
+/** A {@link BrowserPool} of Playwright browsers, as built by {@link playwrightBrowserPool}. */
+export type PlaywrightBrowserPool = BrowserPool<{
+    browserPlugins: [PlaywrightPlugin];
+}, [PlaywrightPlugin]>;
+export interface PlaywrightBrowserPoolOptions extends Omit<BrowserPoolOptions, 'browserPlugins'>, BrowserPoolHooks<ReturnType<PlaywrightPlugin['createController']>, ReturnType<PlaywrightPlugin['createLaunchContext']>, Page> {
+    /** How to launch the browser: which Playwright browser type, proxy, user data dir, ... */
+    launchContext?: PlaywrightLaunchContext;
+    /**
+     * Whether to run the browser in headless mode. Shorthand for `launchContext.launchOptions.headless`.
+     * Defaults to `true`, and can also be set via {@link Configuration}.
+     */
+    headless?: boolean;
+    /** Configuration to read the browser defaults from. Defaults to the global configuration. */
+    configuration?: Configuration;
+}
+export interface RemotePlaywrightBrowserPoolOptions extends Pick<PlaywrightBrowserPoolOptions, 'launchContext' | 'headless' | 'configuration'>, Omit<RemoteBrowserPoolOptions, 'browserPlugins'> {
+}
+/**
+ * Builds a {@link BrowserPool} of Playwright browsers to pass to a {@link PlaywrightCrawler} as its
+ * {@link BrowserCrawlerOptions.browserPool|`browserPool`}.
+ *
+ * It accepts every {@link BrowserPoolOptions|`BrowserPool` option} plus the crawler's own `launchContext` and
+ * `headless`, and derives the browser plugin from them - so a pool built here always matches the crawler it is
+ * given to, and configuring one never means assembling a {@link PlaywrightPlugin} by hand.
+ *
+ * **Example usage:**
+ *
+ * ```javascript
+ * const crawler = new PlaywrightCrawler({
+ *     browserPool: playwrightBrowserPool({
+ *         maxOpenPagesPerBrowser: 1,
+ *         launchContext: { launcher: firefox },
+ *     }),
+ *     requestHandler: async ({ page }) => { ... },
+ * });
+ * ```
+ *
+ * The returned pool is *not* torn down by the crawler, which is what makes it shareable between crawlers.
+ *
+ * @category Browser management
+ */
+export declare function playwrightBrowserPool(options?: PlaywrightBrowserPoolOptions): PlaywrightBrowserPool;
+/**
+ * The {@link RemoteBrowserPool} counterpart of {@link playwrightBrowserPool}: connects to a remote browser
+ * service (Browserbase, Browserless, Steel, ...) with a Playwright plugin derived from `launchContext`.
+ *
+ * A {@link PlaywrightCrawler} accepts the same connection details directly via
+ * {@link BrowserCrawlerOptions.remoteBrowser|`remoteBrowser`}; reach for this factory when you also need to
+ * tune the wrapping pool, or to share one remote pool between crawlers.
+ *
+ * **Example usage:**
+ *
+ * ```javascript
+ * const crawler = new PlaywrightCrawler({
+ *     browserPool: remotePlaywrightBrowserPool({
+ *         endpoint: 'wss://production-sfo.browserless.io?token=xxx',
+ *         maxOpenBrowsers: 2,
+ *         browserPoolOptions: { useFingerprints: false },
+ *     }),
+ *     requestHandler: async ({ page }) => { ... },
+ * });
+ * ```
+ *
+ * @category Browser management
+ */
+export declare function remotePlaywrightBrowserPool(options: RemotePlaywrightBrowserPoolOptions): RemoteBrowserPool<Page>;

package/internals/playwright-browser-pool.js

@@ -0,0 +1,61 @@
+import { PlaywrightLauncher } from './playwright-launcher.js';
+/**
+ * Builds a {@link BrowserPool} of Playwright browsers to pass to a {@link PlaywrightCrawler} as its
+ * {@link BrowserCrawlerOptions.browserPool|`browserPool`}.
+ *
+ * It accepts every {@link BrowserPoolOptions|`BrowserPool` option} plus the crawler's own `launchContext` and
+ * `headless`, and derives the browser plugin from them - so a pool built here always matches the crawler it is
+ * given to, and configuring one never means assembling a {@link PlaywrightPlugin} by hand.
+ *
+ * **Example usage:**
+ *
+ * ```javascript
+ * const crawler = new PlaywrightCrawler({
+ *     browserPool: playwrightBrowserPool({
+ *         maxOpenPagesPerBrowser: 1,
+ *         launchContext: { launcher: firefox },
+ *     }),
+ *     requestHandler: async ({ page }) => { ... },
+ * });
+ * ```
+ *
+ * The returned pool is *not* torn down by the crawler, which is what makes it shareable between crawlers.
+ *
+ * @category Browser management
+ */
+export function playwrightBrowserPool(options = {}) {
+    const { launchContext, headless, configuration, ...poolOptions } = options;
+    return playwrightLauncher(launchContext, headless, configuration).createBrowserPool(poolOptions);
+}
+/**
+ * The {@link RemoteBrowserPool} counterpart of {@link playwrightBrowserPool}: connects to a remote browser
+ * service (Browserbase, Browserless, Steel, ...) with a Playwright plugin derived from `launchContext`.
+ *
+ * A {@link PlaywrightCrawler} accepts the same connection details directly via
+ * {@link BrowserCrawlerOptions.remoteBrowser|`remoteBrowser`}; reach for this factory when you also need to
+ * tune the wrapping pool, or to share one remote pool between crawlers.
+ *
+ * **Example usage:**
+ *
+ * ```javascript
+ * const crawler = new PlaywrightCrawler({
+ *     browserPool: remotePlaywrightBrowserPool({
+ *         endpoint: 'wss://production-sfo.browserless.io?token=xxx',
+ *         maxOpenBrowsers: 2,
+ *         browserPoolOptions: { useFingerprints: false },
+ *     }),
+ *     requestHandler: async ({ page }) => { ... },
+ * });
+ * ```
+ *
+ * @category Browser management
+ */
+export function remotePlaywrightBrowserPool(options) {
+    const { launchContext, headless, configuration, ...remoteOptions } = options;
+    return playwrightLauncher(launchContext, headless, configuration).createRemoteBrowserPool(remoteOptions);
+}
+function playwrightLauncher(launchContext = {}, headless, configuration) {
+    return new PlaywrightLauncher(headless == null
+        ? launchContext
+        : { ...launchContext, launchOptions: { ...launchContext.launchOptions, headless } }, configuration);
+}