@crawlee/core 4.0.0-beta.120 → 4.0.0-beta.121

package/crawlers/error_snapshotter.d.ts

@@ -46,7 +46,7 @@ export declare class ErrorSnapshotter {
      */
     contextCaptureSnapshot(context: BrowserCrawlingContext, fileName: string): Promise<SnapshotResult | undefined>;
     /**
-     * Save the HTML snapshot of the page, and return the fileName with the extension.
+     * Save the HTML snapshot of the page, and return the key it was stored under.
      */
     saveHTMLSnapshot(html: string, keyValueStore: Pick<KeyValueStore, 'setValue'>, fileName: string): Promise<string | undefined>;
     /**

package/crawlers/error_snapshotter.js

@@ -77,12 +77,14 @@ export class ErrorSnapshotter {
         }
     }
     /**
-     * Save the HTML snapshot of the page, and return the fileName with the extension.
+     * Save the HTML snapshot of the page, and return the key it was stored under.
      */
     async saveHTMLSnapshot(html, keyValueStore, fileName) {
         try {
             await keyValueStore.setValue(fileName, html, { contentType: 'text/html' });
-            return `${fileName}.html`;
+            // The record key is `fileName` - returning it with an `.html` suffix (as v3 did,
+            // where local storage put the extension in the key) would break `getPublicUrl`.
+            return fileName;
         }
         catch {
             return undefined;

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@crawlee/core",
-    "version": "4.0.0-beta.120",
+    "version": "4.0.0-beta.121",
     "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"
@@ -52,9 +52,9 @@
         "@apify/log": "^2.5.18",
         "@apify/timeout": "^0.4.4",
         "@apify/utilities": "^2.15.5",
-        "@crawlee/fs-storage": "4.0.0-beta.120",
-        "@crawlee/types": "4.0.0-beta.120",
-        "@crawlee/utils": "4.0.0-beta.120",
+        "@crawlee/fs-storage": "4.0.0-beta.121",
+        "@crawlee/types": "4.0.0-beta.121",
+        "@crawlee/utils": "4.0.0-beta.121",
         "@sapphire/async-queue": "^1.5.5",
         "@sapphire/shapeshift": "^4.0.0",
         "@vladfrangu/async_event_emitter": "^2.4.6",
@@ -78,5 +78,5 @@
             }
         }
     },
-    "gitHead": "da5d427c4c1c9dcaea95b151e9c4c7310100885d"
+    "gitHead": "5027317de626f5ba6de5047ae9341a898258cc5a"
 }

package/request.js

@@ -167,7 +167,10 @@ class CrawleeRequest {
         if (label) {
             userData.label = label;
         }
-        this.#userData = { __crawlee: {}, ...userData };
+        // Read `__crawlee` explicitly - on a `userData` coming from another Request instance the
+        // bag is non-enumerable, so the spread alone would silently drop the internal state
+        // (e.g. `skipNavigation`) when a request is re-wrapped after a storage round trip.
+        this.#userData = { __crawlee: userData.__crawlee ?? {}, ...userData };
         // `userData` must stay an enumerable own accessor — serialization in the storages relies on it
         Object.defineProperties(this, {
             userData: {

package/service_locator.d.ts

@@ -62,6 +62,17 @@ interface ServiceLocatorInterface {
      * Get the storage instance manager (shared across all storage types).
      */
     getStorageInstanceManager(): StorageInstanceManager;
+    /**
+     * Returns the currently set services without triggering the implicit creation of defaults.
+     * Used to inherit already-materialized services into crawler-scoped service locators.
+     * @internal
+     */
+    getServicesIfSet(): {
+        configuration?: Configuration;
+        eventManager?: EventManager;
+        storageBackend?: StorageBackend;
+        logger?: CrawleeLogger;
+    };
     /**
      * Resets the service locator to its initial state.
      * Used mainly for testing purposes.
@@ -111,6 +122,13 @@ export declare class ServiceLocator implements ServiceLocatorInterface {
      * @param logger Optional logger instance to use
      */
     constructor(configuration?: Configuration, eventManager?: EventManager, storageBackend?: StorageBackend, logger?: CrawleeLogger);
+    /** @internal */
+    getServicesIfSet(): {
+        configuration?: Configuration;
+        eventManager?: EventManager;
+        storageBackend?: StorageBackend;
+        logger?: CrawleeLogger;
+    };
     getConfiguration(): Configuration;
     setConfiguration(configuration: Configuration): void;
     getEventManager(): EventManager;

package/service_locator.js

@@ -63,6 +63,15 @@ export class ServiceLocator {
         this.#storageBackend = storageBackend;
         this.#logger = logger;
     }
+    /** @internal */
+    getServicesIfSet() {
+        return {
+            configuration: this.#configuration,
+            eventManager: this.#eventManager,
+            storageBackend: this.#storageBackend,
+            logger: this.#logger,
+        };
+    }
     getConfiguration() {
         if (!this.#configuration) {
             this.getLogger().debug('No configuration set, implicitly creating and using default Configuration.');