@crawlee/browser-pool 4.0.0-beta.26 → 4.0.0-beta.260

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (86) hide show
  1. package/README.md +15 -15
  2. package/abstract-classes/browser-controller.d.ts +100 -29
  3. package/abstract-classes/browser-controller.js +70 -24
  4. package/abstract-classes/browser-plugin.d.ts +25 -22
  5. package/abstract-classes/browser-plugin.js +21 -13
  6. package/anonymize-proxy.d.ts +6 -2
  7. package/anonymize-proxy.js +9 -5
  8. package/browser-pool.d.ts +110 -51
  9. package/browser-pool.js +318 -147
  10. package/events.d.ts +2 -4
  11. package/events.js +0 -2
  12. package/fingerprinting/hooks.d.ts +0 -1
  13. package/fingerprinting/hooks.js +26 -10
  14. package/fingerprinting/types.d.ts +3 -34
  15. package/fingerprinting/types.js +0 -2
  16. package/fingerprinting/utils.d.ts +0 -1
  17. package/fingerprinting/utils.js +5 -6
  18. package/index.d.ts +15 -7
  19. package/index.js +6 -3
  20. package/launch-context.d.ts +15 -5
  21. package/launch-context.js +15 -10
  22. package/package.json +15 -15
  23. package/playwright/playwright-browser.d.ts +3 -6
  24. package/playwright/playwright-browser.js +17 -17
  25. package/playwright/playwright-controller.d.ts +1 -1
  26. package/playwright/playwright-controller.js +10 -6
  27. package/playwright/playwright-plugin.d.ts +3 -7
  28. package/playwright/playwright-plugin.js +18 -17
  29. package/playwright/remote-playwright-plugin.d.ts +29 -0
  30. package/playwright/remote-playwright-plugin.js +38 -0
  31. package/proxy-server.d.ts +0 -1
  32. package/proxy-server.js +0 -1
  33. package/puppeteer/puppeteer-controller.d.ts +1 -1
  34. package/puppeteer/puppeteer-controller.js +29 -20
  35. package/puppeteer/puppeteer-plugin.d.ts +12 -4
  36. package/puppeteer/puppeteer-plugin.js +75 -48
  37. package/puppeteer/remote-puppeteer-plugin.d.ts +22 -0
  38. package/puppeteer/remote-puppeteer-plugin.js +31 -0
  39. package/remote-browser-plugin.d.ts +49 -0
  40. package/remote-browser-plugin.js +63 -0
  41. package/remote-browser-provider.d.ts +81 -0
  42. package/remote-browser-provider.js +65 -0
  43. package/utils.d.ts +7 -1
  44. package/utils.js +19 -1
  45. package/abstract-classes/browser-controller.d.ts.map +0 -1
  46. package/abstract-classes/browser-controller.js.map +0 -1
  47. package/abstract-classes/browser-plugin.d.ts.map +0 -1
  48. package/abstract-classes/browser-plugin.js.map +0 -1
  49. package/anonymize-proxy.d.ts.map +0 -1
  50. package/anonymize-proxy.js.map +0 -1
  51. package/browser-pool.d.ts.map +0 -1
  52. package/browser-pool.js.map +0 -1
  53. package/container-proxy-server.d.ts +0 -12
  54. package/container-proxy-server.d.ts.map +0 -1
  55. package/container-proxy-server.js +0 -43
  56. package/container-proxy-server.js.map +0 -1
  57. package/events.d.ts.map +0 -1
  58. package/events.js.map +0 -1
  59. package/fingerprinting/hooks.d.ts.map +0 -1
  60. package/fingerprinting/hooks.js.map +0 -1
  61. package/fingerprinting/types.d.ts.map +0 -1
  62. package/fingerprinting/types.js.map +0 -1
  63. package/fingerprinting/utils.d.ts.map +0 -1
  64. package/fingerprinting/utils.js.map +0 -1
  65. package/index.d.ts.map +0 -1
  66. package/index.js.map +0 -1
  67. package/launch-context.d.ts.map +0 -1
  68. package/launch-context.js.map +0 -1
  69. package/logger.d.ts +0 -3
  70. package/logger.d.ts.map +0 -1
  71. package/logger.js +0 -5
  72. package/logger.js.map +0 -1
  73. package/playwright/playwright-browser.d.ts.map +0 -1
  74. package/playwright/playwright-browser.js.map +0 -1
  75. package/playwright/playwright-controller.d.ts.map +0 -1
  76. package/playwright/playwright-controller.js.map +0 -1
  77. package/playwright/playwright-plugin.d.ts.map +0 -1
  78. package/playwright/playwright-plugin.js.map +0 -1
  79. package/proxy-server.d.ts.map +0 -1
  80. package/proxy-server.js.map +0 -1
  81. package/puppeteer/puppeteer-controller.d.ts.map +0 -1
  82. package/puppeteer/puppeteer-controller.js.map +0 -1
  83. package/puppeteer/puppeteer-plugin.d.ts.map +0 -1
  84. package/puppeteer/puppeteer-plugin.js.map +0 -1
  85. package/utils.d.ts.map +0 -1
  86. package/utils.js.map +0 -1
@@ -1,15 +1,20 @@
1
- import { anonymizeProxy, closeAnonymizedProxy } from 'proxy-chain';
2
- export const anonymizeProxySugar = async (proxyUrl, username, password) => {
1
+ /** @internal */
2
+ export const anonymizeProxySugar = async (proxyUrl, username, password, options) => {
3
3
  if (proxyUrl) {
4
4
  const url = new URL(proxyUrl);
5
5
  if (username || password) {
6
6
  url.username = username ?? '';
7
7
  url.password = password ?? '';
8
8
  }
9
- if (url.username || url.password) {
9
+ if (url.username || url.password || options?.ignoreProxyCertificate) {
10
10
  // trim off trailing slash if it's present
11
11
  const proxyUrlString = url.href.endsWith('/') ? url.href.slice(0, -1) : url.href;
12
- const anonymized = await anonymizeProxy(proxyUrlString);
12
+ const { anonymizeProxy, closeAnonymizedProxy } = await import('proxy-chain');
13
+ const anonymized = await anonymizeProxy({
14
+ url: proxyUrlString,
15
+ port: 0,
16
+ ignoreProxyCertificate: options?.ignoreProxyCertificate ?? false,
17
+ });
13
18
  return [
14
19
  anonymized,
15
20
  async () => {
@@ -21,4 +26,3 @@ export const anonymizeProxySugar = async (proxyUrl, username, password) => {
21
26
  }
22
27
  return [undefined, async () => { }];
23
28
  };
24
- //# sourceMappingURL=anonymize-proxy.js.map
package/browser-pool.d.ts CHANGED
@@ -1,3 +1,4 @@
1
+ import type { IBrowserPool, NewPageOptions, PageState } from '@crawlee/types';
1
2
  import type { BrowserFingerprintWithHeaders } from 'fingerprint-generator';
2
3
  import { FingerprintGenerator } from 'fingerprint-generator';
3
4
  import { FingerprintInjector } from 'fingerprint-injector';
@@ -53,6 +54,16 @@ export interface BrowserPoolOptions<Plugin extends BrowserPlugin = BrowserPlugin
53
54
  * @default 20
54
55
  */
55
56
  maxOpenPagesPerBrowser?: number;
57
+ /**
58
+ * Maximum number of browser processes alive at once. A browser counts from the moment its launch starts until it
59
+ * actually closes — a retired browser that is still finishing its last pages still occupies a slot. Once the
60
+ * limit is reached, {@link BrowserPool.newPage|newPage} waits for a slot (a browser closing, or an open
61
+ * browser with page capacity to spare) instead of launching another. Set it to a remote browser service's
62
+ * concurrent-session limit to avoid `429` errors.
63
+ *
64
+ * @default Infinity
65
+ */
66
+ maxOpenBrowsers?: number;
56
67
  /**
57
68
  * Browsers tend to get bloated after processing a lot of pages. This option
58
69
  * configures the maximum number of processed pages after which the browser will
@@ -246,38 +257,37 @@ export interface BrowserPoolHooks<BC extends BrowserController, LC extends Launc
246
257
  * });
247
258
  * ```
248
259
  */
249
- export declare class BrowserPool<Options extends BrowserPoolOptions = BrowserPoolOptions, BrowserPlugins extends BrowserPlugin[] = InferBrowserPluginArray<Options['browserPlugins']>, BrowserControllerReturn extends BrowserController = ReturnType<BrowserPlugins[number]['createController']>, LaunchContextReturn extends LaunchContext = ReturnType<BrowserPlugins[number]['createLaunchContext']>, PageOptions = Parameters<BrowserControllerReturn['newPage']>[0], PageReturn extends UnwrapPromise<ReturnType<BrowserControllerReturn['newPage']>> = UnwrapPromise<ReturnType<BrowserControllerReturn['newPage']>>> extends TypedEmitter<BrowserPoolEvents<BrowserControllerReturn, PageReturn>> {
260
+ export declare class BrowserPool<Options extends BrowserPoolOptions = BrowserPoolOptions, BrowserPlugins extends BrowserPlugin[] = InferBrowserPluginArray<Options['browserPlugins']>, BrowserControllerReturn extends BrowserController = ReturnType<BrowserPlugins[number]['createController']>, LaunchContextReturn extends LaunchContext = ReturnType<BrowserPlugins[number]['createLaunchContext']>, PageOptions = Parameters<BrowserControllerReturn['newPage']>[0], PageReturn extends UnwrapPromise<ReturnType<BrowserControllerReturn['newPage']>> = UnwrapPromise<ReturnType<BrowserControllerReturn['newPage']>>> extends TypedEmitter<BrowserPoolEvents<BrowserControllerReturn, PageReturn>> implements IBrowserPool<PageReturn> {
261
+ #private;
250
262
  browserPlugins: BrowserPlugins;
251
- maxOpenPagesPerBrowser: number;
252
- retireBrowserAfterPageCount: number;
253
- operationTimeoutMillis: number;
254
- closeInactiveBrowserAfterMillis: number;
255
- useFingerprints?: boolean;
263
+ /** @internal */
256
264
  fingerprintOptions: FingerprintOptions;
257
- preLaunchHooks: PreLaunchHook<LaunchContextReturn>[];
258
- postLaunchHooks: PostLaunchHook<BrowserControllerReturn>[];
259
- prePageCreateHooks: PrePageCreateHook<BrowserControllerReturn, PageOptions>[];
260
- postPageCreateHooks: PostPageCreateHook<BrowserControllerReturn, PageReturn>[];
261
- prePageCloseHooks: PrePageCloseHook<BrowserControllerReturn, PageReturn>[];
262
- postPageCloseHooks: PostPageCloseHook<BrowserControllerReturn>[];
263
- pageCounter: number;
264
- pages: Map<string, PageReturn>;
265
- pageIds: WeakMap<PageReturn, string>;
266
- startingBrowserControllers: Set<BrowserControllerReturn>;
267
- activeBrowserControllers: Set<BrowserControllerReturn>;
268
- retiredBrowserControllers: Set<BrowserControllerReturn>;
269
- pageToBrowserController: WeakMap<PageReturn, BrowserControllerReturn>;
265
+ /** @internal */
270
266
  fingerprintInjector?: FingerprintInjector;
271
267
  fingerprintGenerator?: FingerprintGenerator;
268
+ /** @internal */
272
269
  fingerprintCache?: QuickLRU<string, BrowserFingerprintWithHeaders>;
270
+ private pages;
271
+ private activeBrowserControllers;
272
+ private retiredBrowserControllers;
273
273
  private browserKillerInterval?;
274
- private browserRetireInterval?;
275
- private limiter;
276
274
  constructor(options: Options & BrowserPoolHooks<BrowserControllerReturn, LaunchContextReturn, PageReturn>);
277
275
  /**
278
276
  * Opens a new page in one of the running browsers or launches
279
277
  * a new browser and opens a page there, if no browsers are active,
280
278
  * or their page limits have been exceeded.
279
+ *
280
+ * **Session injection (best-effort):** When a {@link NewPageOptions.session|session} is
281
+ * provided, this implementation uses it as a cache key for browser fingerprints (when
282
+ * fingerprinting is enabled) and reads
283
+ * {@link ProxyInfo.url|session.proxyInfo.url} /
284
+ * {@link ProxyInfo.ignoreTlsErrors|session.proxyInfo.ignoreTlsErrors} as defaults
285
+ * for `proxyUrl` and `ignoreTlsErrors` respectively. Explicit `proxyUrl` /
286
+ * `ignoreTlsErrors` values in the options take precedence.
287
+ *
288
+ * Beyond fingerprint caching and proxy configuration, no other session
289
+ * properties are consumed — cookie and header injection remain the
290
+ * crawler's responsibility.
281
291
  */
282
292
  newPage(options?: BrowserPoolNewPageOptions<PageOptions, BrowserPlugins[number]>): Promise<PageReturn>;
283
293
  /**
@@ -335,7 +345,6 @@ export declare class BrowserPool<Options extends BrowserPoolOptions = BrowserPoo
335
345
  * until it's closed.
336
346
  */
337
347
  getPageId(page: PageReturn): string | undefined;
338
- private _createPageForBrowser;
339
348
  /**
340
349
  * Removes a browser controller from the pool. The underlying
341
350
  * browser will be closed after all its pages are closed.
@@ -347,6 +356,44 @@ export declare class BrowserPool<Options extends BrowserPoolOptions = BrowserPoo
347
356
  * closed after all its pages are closed.
348
357
  */
349
358
  retireBrowserByPage(page: PageReturn): void;
359
+ /**
360
+ * Releases a page back to the pool. The page is closed and, if the
361
+ * optional `error` is a {@link SessionError} (including
362
+ * {@link SessionRetiredError}), the browser controller that served
363
+ * the page is retired so that the finished session's state (cookies,
364
+ * storage, etc.) cannot leak into future sessions.
365
+ *
366
+ * This is the primary way the crawler should return pages to the pool.
367
+ *
368
+ * @param page The page to release.
369
+ * @param options.error The error that caused the page to be released, if any.
370
+ */
371
+ closePage(page: PageReturn, options?: {
372
+ error?: Error;
373
+ }): Promise<void>;
374
+ /**
375
+ * Extracts the relevant state (currently just cookies) from a page via its
376
+ * owning {@link BrowserController}. Returns empty state when the page is
377
+ * no longer associated with a controller.
378
+ *
379
+ * As with {@link BrowserPool.injectPageState}, cookies are isolated per
380
+ * page only when the pool is configured with `useIncognitoPages: true`.
381
+ * With the default `useIncognitoPages: false`, the extracted cookies
382
+ * include those set by any sibling page sharing the same browser.
383
+ */
384
+ extractPageState(page: PageReturn): Promise<PageState>;
385
+ /**
386
+ * Injects state into a page via its owning {@link BrowserController}.
387
+ *
388
+ * No-op when the page is no longer associated with a controller.
389
+ *
390
+ * Note that cookies are isolated per page only when the pool is configured
391
+ * with `useIncognitoPages: true` — each page then gets its own browser
392
+ * context. With the default `useIncognitoPages: false`, all pages in a
393
+ * browser share a single context, so injected cookies are visible to every
394
+ * page served by that browser.
395
+ */
396
+ injectPageState(page: PageReturn, state: PageState): Promise<void>;
350
397
  /**
351
398
  * Removes all active browsers from the pool. The browsers will be
352
399
  * closed after all their pages are closed.
@@ -357,32 +404,53 @@ export declare class BrowserPool<Options extends BrowserPoolOptions = BrowserPoo
357
404
  * @return {Promise<void>}
358
405
  */
359
406
  closeAllBrowsers(): Promise<void>;
407
+ [Symbol.asyncDispose](): Promise<void>;
360
408
  /**
361
- * Closes all managed browsers and tears down the pool.
409
+ * Closes every managed browser and empties the pool, which stays usable afterwards — a crawler releases its
410
+ * browsers when a run ends and may start another run on the same pool.
411
+ */
412
+ releaseAllBrowsers(): Promise<void>;
413
+ /**
414
+ * Closes all managed browsers and tears the pool down for good: its intervals are cleared without being
415
+ * re-armed and its listeners are dropped, so it cannot be used again.
362
416
  */
363
417
  destroy(): Promise<void>;
364
- private _teardown;
365
- private _getAllBrowserControllers;
366
- private _launchBrowser;
367
- /**
368
- * Picks plugins round robin.
369
- * @private
370
- */
371
- private _pickBrowserPlugin;
372
- private _pickBrowserWithFreeCapacity;
373
- private _closeInactiveRetiredBrowsers;
374
- private _overridePageClose;
375
- private _executeHooks;
376
- private _closeRetiredBrowserWithNoPages;
377
- private _initializeFingerprinting;
378
- private _addFingerprintHooks;
418
+ private closeInactiveRetiredBrowsers;
419
+ /**
420
+ * Returns `true` if the pool can accept a new browser launch without exceeding `maxOpenBrowsers`.
421
+ * Counts starting, active, and retired browsers.
422
+ *
423
+ * @internal
424
+ */
425
+ hasFreeBrowserSlot(): boolean;
379
426
  }
380
- export interface BrowserPoolNewPageOptions<PageOptions, BP extends BrowserPlugin> {
427
+ export interface BrowserPoolNewPageOptions<PageOptions, BP extends BrowserPlugin> extends NewPageOptions {
428
+ /**
429
+ * The proxy URL the pool uses internally to route the page: it keys browser
430
+ * reuse (with `browserPerProxy`, only a browser already on this proxy is
431
+ * reused), configures the launched browser, and is applied to incognito
432
+ * pages. When omitted, it is derived from the
433
+ * {@link NewPageOptions.session|session}'s `proxyInfo`; an explicit value
434
+ * here takes precedence.
435
+ *
436
+ * This is an implementation detail of the built-in `BrowserPool`'s proxy
437
+ * handling and is intentionally not part of the {@link IBrowserPool}
438
+ * contract — through that interface the proxy is supplied via the session.
439
+ */
440
+ proxyUrl?: string;
381
441
  /**
382
- * Assign a custom ID to the page. If you don't a random string ID
383
- * will be generated.
442
+ * Disable TLS certificate verification for MITM proxies. Applied both when
443
+ * launching a new browser and when creating a page in an existing one. When
444
+ * omitted, it is derived from the
445
+ * {@link NewPageOptions.session|session}'s `proxyInfo`; an explicit value
446
+ * here takes precedence.
447
+ *
448
+ * This is an implementation detail of the built-in `BrowserPool` and is
449
+ * intentionally not part of the {@link IBrowserPool} contract — through
450
+ * that interface, configure it via the session's `proxyInfo` or through the
451
+ * browser's `launchOptions`.
384
452
  */
385
- id?: string;
453
+ ignoreTlsErrors?: boolean;
386
454
  /**
387
455
  * Some libraries (Playwright) allow you to open new pages with specific
388
456
  * options. Use this property to set those options.
@@ -397,14 +465,6 @@ export interface BrowserPoolNewPageOptions<PageOptions, BP extends BrowserPlugin
397
465
  * see the `newPageInNewBrowser` function.
398
466
  */
399
467
  browserPlugin?: BP;
400
- /**
401
- * Proxy URL.
402
- */
403
- proxyUrl?: string;
404
- /**
405
- * Proxy tier.
406
- */
407
- proxyTier?: number;
408
468
  }
409
469
  export interface BrowserPoolNewPageInNewBrowserOptions<PageOptions, BP extends BrowserPlugin> {
410
470
  /**
@@ -435,4 +495,3 @@ export interface BrowserPoolNewPageInNewBrowserOptions<PageOptions, BP extends B
435
495
  */
436
496
  launchOptions?: BP['launchOptions'];
437
497
  }
438
- //# sourceMappingURL=browser-pool.d.ts.map