@crawlee/basic 4.0.0-beta.115 → 4.0.0-beta.117

package/internals/basic-crawler.js

@@ -1,6 +1,6 @@
 import { mkdir, writeFile } from 'node:fs/promises';
 import { dirname } from 'node:path';
-import { AutoscaledPool, bindMethodsToServiceLocator, BLOCKED_STATUS_CODES, ConcurrencySystem, ContextPipeline, ContextPipelineCleanupError, ContextPipelineInitializationError, ContextPipelineInterruptedError, createStorageTransaction, CriticalError, currentStorageTransaction, Dataset, enqueueLinks, EnqueueStrategy, getObjectType, KeyValueStore, log, LogLevel, mergeCookies, MissingSessionError, NavigationSkippedError, NonRetryableError, OwnedOrInjected, purgeDefaultStorages, RequestHandlerError, parseRetryAfterHeader, RequestThrottledError, RequestManagerTandem, RequestQueue, RequestState, RetryRequestError, supportsDomainThrottling, Router, ServiceLocator, serviceLocator, Session, SessionError, SessionPool, Statistics, validateUserData, validators, withDirectStorageAccess, } from '@crawlee/core';
+import { AutoscaledPool, bindMethodsToServiceLocator, BLOCKED_STATUS_CODES, ConcurrencySystem, ContextPipeline, ContextPipelineCleanupError, ContextPipelineInitializationError, ContextPipelineInterruptedError, createStorageTransaction, CriticalError, currentStorageTransaction, Dataset, enqueueLinks, EnqueueStrategy, EventType, getObjectType, KeyValueStore, log, LogLevel, mergeCookies, MissingSessionError, NavigationSkippedError, NonRetryableError, OwnedOrInjected, purgeDefaultStorages, RequestHandlerError, parseRetryAfterHeader, RequestThrottledError, RequestManagerTandem, RequestQueue, RequestState, RetryRequestError, supportsDomainThrottling, Router, ServiceLocator, serviceLocator, Session, SessionError, SessionPool, Statistics, validateUserData, validators, withDirectStorageAccess, } from '@crawlee/core';
 import { FetchHttpClient } from '@crawlee/http-client';
 import { isAsyncIterable, isIterable, ROTATE_PROXY_ERRORS } from '@crawlee/utils/internal';
 import { RobotsTxtFile } from '@crawlee/utils';
@@ -662,7 +662,7 @@ export class BasicCrawler {
         // subscribe to `EventType.STATUS_MESSAGE` and propagate it to their status-reporting backend.
         // Setting the status message is not a storage concern, so we intentionally don't route it
         // through the storage client anymore.
-        serviceLocator.getEventManager().emit("statusMessage" /* EventType.STATUS_MESSAGE */, {
+        serviceLocator.getEventManager().emit(EventType.STATUS_MESSAGE, {
             crawlerId: this.identity.id,
             message,
             isStatusMessageTerminal: options.isStatusMessageTerminal,
@@ -776,8 +776,8 @@ export class BasicCrawler {
         const boundPauseOnMigration = this.pauseOnMigration.bind(this);
         process.once('SIGINT', sigintHandler);
         const eventManager = serviceLocator.getEventManager();
-        eventManager.on("migrating" /* EventType.MIGRATING */, boundPauseOnMigration);
-        eventManager.on("aborting" /* EventType.ABORTING */, boundPauseOnMigration);
+        eventManager.on(EventType.MIGRATING, boundPauseOnMigration);
+        eventManager.on(EventType.ABORTING, boundPauseOnMigration);
         let stats = {};
         try {
             await this.#autoscaledPool.run();
@@ -786,8 +786,8 @@ export class BasicCrawler {
             await this.teardown();
             await this.stats.stopCapturing();
             process.off('SIGINT', sigintHandler);
-            eventManager.off("migrating" /* EventType.MIGRATING */, boundPauseOnMigration);
-            eventManager.off("aborting" /* EventType.ABORTING */, boundPauseOnMigration);
+            eventManager.off(EventType.MIGRATING, boundPauseOnMigration);
+            eventManager.off(EventType.ABORTING, boundPauseOnMigration);
             const finalStats = this.stats.calculate();
             stats = {
                 requestsFinished: this.stats.state.requestsFinished,
@@ -1668,7 +1668,7 @@ export class BasicCrawler {
      * To stop the crawler gracefully (waiting for all running requests to finish), use {@link BasicCrawler.stop|`crawler.stop()`} instead.
      */
     async teardown() {
-        serviceLocator.getEventManager().emit("persistState" /* EventType.PERSIST_STATE */, { isMigrating: false });
+        serviceLocator.getEventManager().emit(EventType.PERSIST_STATE, { isMigrating: false });
         if (this.#closeEvents) {
             await serviceLocator.getEventManager().close();
         }

package/internals/send-request.js

@@ -27,6 +27,7 @@ export function createSendRequest(httpClient, originRequest, session) {
             cookieJar: overrideOptions?.cookieJar ?? session.cookieJar,
             timeoutMillis: overrideOptions.timeoutMillis,
             signal: overrideOptions.signal,
+            ignoreTlsErrors: overrideOptions.ignoreTlsErrors,
         });
     };
 }

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@crawlee/basic",
-    "version": "4.0.0-beta.115",
+    "version": "4.0.0-beta.117",
     "description": "The scalable web crawling and scraping library for JavaScript/Node.js. Enables development of data extraction and web automation jobs (not only) with headless Chrome and Puppeteer.",
     "engines": {
         "node": ">=22.0.0"
@@ -42,10 +42,10 @@
         "@apify/datastructures": "^2.0.0",
         "@apify/timeout": "^0.4.4",
         "@apify/utilities": "^2.15.5",
-        "@crawlee/core": "4.0.0-beta.115",
-        "@crawlee/http-client": "4.0.0-beta.115",
-        "@crawlee/types": "4.0.0-beta.115",
-        "@crawlee/utils": "4.0.0-beta.115",
+        "@crawlee/core": "4.0.0-beta.117",
+        "@crawlee/http-client": "4.0.0-beta.117",
+        "@crawlee/types": "4.0.0-beta.117",
+        "@crawlee/utils": "4.0.0-beta.117",
         "csv-stringify": "^6.5.2",
         "ow": "^2.0.0",
         "tldts": "^7.0.6",
@@ -53,7 +53,7 @@
         "type-fest": "^4.41.0"
     },
     "optionalDependencies": {
-        "@crawlee/impit-client": "^4.0.0-beta.115"
+        "@crawlee/impit-client": "^4.0.0-beta.117"
     },
     "lerna": {
         "command": {
@@ -62,5 +62,5 @@
             }
         }
     },
-    "gitHead": "9bb6f54b10a88a709f0461486601a98365a90403"
+    "gitHead": "7c9c8c83da02da0444130cb95368045d30881c3b"
 }