@crawlee/core 4.0.0-beta.83 → 4.0.0-beta.85

package/debug.d.ts

@@ -0,0 +1,36 @@
+import type { IncomingMessage } from 'node:http';
+import type { Dictionary } from '@crawlee/types';
+import type { Request } from './request.js';
+interface BrowserResponseLike {
+    status(): number;
+}
+/**
+ * Creates a standardized debug info from request and response. This info is usually added to dataset under the hidden `#debug` field.
+ *
+ * @param request [Request](https://sdk.apify.com/docs/api/request) object.
+ * @param [response]
+ *   Puppeteer [`Response`](https://pptr.dev/#?product=Puppeteer&version=v1.11.0&show=api-class-response)
+ *   or NodeJS [`http.IncomingMessage`](https://nodejs.org/api/http.html#http_class_http_serverresponse).
+ * @param [additionalFields] Object containing additional fields to be added.
+ *
+ * @internal
+ */
+export declare function createRequestDebugInfo(request: Request, response?: IncomingMessage | Partial<BrowserResponseLike>, additionalFields?: Dictionary): Dictionary;
+/**
+ * Returns a human-readable label for an unknown value,
+ * suitable for embedding in error messages and log output.
+ *
+ * Returns `constructor.name` when available (e.g. `"Configuration"`, `"Number"`),
+ * otherwise falls back to `util.inspect` (e.g. for `null`, `undefined`).
+ *
+ * @internal
+ */
+export declare function inspectValue(value: unknown): string;
+/**
+ * Returns the type of a value as a lowercase string, with `Date`, `Buffer` and `RegExp` reported
+ * by their constructor name. Used for building validation error messages.
+ *
+ * @internal
+ */
+export declare function getObjectType(value: unknown): string;
+export {};

package/debug.js

@@ -0,0 +1,70 @@
+import { inspect } from 'node:util';
+import ow from 'ow';
+/**
+ * Creates a standardized debug info from request and response. This info is usually added to dataset under the hidden `#debug` field.
+ *
+ * @param request [Request](https://sdk.apify.com/docs/api/request) object.
+ * @param [response]
+ *   Puppeteer [`Response`](https://pptr.dev/#?product=Puppeteer&version=v1.11.0&show=api-class-response)
+ *   or NodeJS [`http.IncomingMessage`](https://nodejs.org/api/http.html#http_class_http_serverresponse).
+ * @param [additionalFields] Object containing additional fields to be added.
+ *
+ * @internal
+ */
+export function createRequestDebugInfo(request, response = {}, additionalFields = {}) {
+    ow(request, ow.object);
+    ow(response, ow.object);
+    ow(additionalFields, ow.object);
+    return {
+        requestId: request.id,
+        url: request.url,
+        loadedUrl: request.loadedUrl,
+        method: request.method,
+        retryCount: request.retryCount,
+        errorMessages: request.errorMessages,
+        // Puppeteer response has .status() function and NodeJS response, statusCode property.
+        statusCode: 'status' in response && response.status instanceof Function
+            ? response.status()
+            : response.statusCode,
+        ...additionalFields,
+    };
+}
+/**
+ * Returns a human-readable label for an unknown value,
+ * suitable for embedding in error messages and log output.
+ *
+ * Returns `constructor.name` when available (e.g. `"Configuration"`, `"Number"`),
+ * otherwise falls back to `util.inspect` (e.g. for `null`, `undefined`).
+ *
+ * @internal
+ */
+export function inspectValue(value) {
+    if (typeof value === 'object' && value !== null && value.constructor?.name) {
+        return value.constructor.name;
+    }
+    return inspect(value, {
+        depth: 0,
+        compact: true,
+        maxStringLength: 64,
+        breakLength: Infinity,
+        colors: false,
+    });
+}
+/**
+ * Returns the type of a value as a lowercase string, with `Date`, `Buffer` and `RegExp` reported
+ * by their constructor name. Used for building validation error messages.
+ *
+ * @internal
+ */
+export function getObjectType(value) {
+    const simple = typeof value;
+    if (['string', 'number', 'boolean', 'bigint'].includes(simple)) {
+        return simple;
+    }
+    const objectType = Object.prototype.toString.call(value);
+    const type = /\[object (\w+)]/.exec(objectType)[1];
+    if (type === 'Uint8Array') {
+        return 'Buffer';
+    }
+    return ['Date', 'Buffer', 'RegExp'].includes(type) ? type : type.toLowerCase();
+}

package/errors.js

@@ -1,4 +1,4 @@
-import { inspectValue } from '@crawlee/utils';
+import { inspectValue } from './debug.js';
 /**
  * Errors of `NonRetryableError` type will never be retried by the crawler.
  */

package/index.d.ts

@@ -1,3 +1,4 @@
+export * from './debug.js';
 export * from './errors.js';
 export * from './autoscaling/index.js';
 export * from './configuration.js';

package/index.js

@@ -1,3 +1,4 @@
+export * from './debug.js';
 export * from './errors.js';
 export * from './autoscaling/index.js';
 export * from './configuration.js';

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@crawlee/core",
-    "version": "4.0.0-beta.83",
+    "version": "4.0.0-beta.85",
     "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"
@@ -53,9 +53,9 @@
         "@apify/pseudo_url": "^2.0.59",
         "@apify/timeout": "^0.3.2",
         "@apify/utilities": "^2.15.5",
-        "@crawlee/fs-storage": "4.0.0-beta.83",
-        "@crawlee/types": "4.0.0-beta.83",
-        "@crawlee/utils": "4.0.0-beta.83",
+        "@crawlee/fs-storage": "4.0.0-beta.85",
+        "@crawlee/types": "4.0.0-beta.85",
+        "@crawlee/utils": "4.0.0-beta.85",
         "@sapphire/async-queue": "^1.5.5",
         "@sapphire/shapeshift": "^4.0.0",
         "@vladfrangu/async_event_emitter": "^2.4.6",
@@ -79,5 +79,5 @@
             }
         }
     },
-    "gitHead": "5f25b90c4914d7e097732cf761e440aeba799137"
+    "gitHead": "8d6d3cd9da3dc9cb0b7822b28eb317019b6806bc"
 }

package/storages/request_queue.js

@@ -1,8 +1,9 @@
 import { inspect } from 'node:util';
-import { chunkedAsyncIterable, downloadListOfUrls, getObjectType, isAsyncIterable, isIterable, peekableAsyncIterable, sleep, } from '@crawlee/utils';
+import { chunkedAsyncIterable, downloadListOfUrls, isAsyncIterable, isIterable, peekableAsyncIterable, sleep, } from '@crawlee/utils';
 import ow from 'ow';
 import { LruCache } from '@apify/datastructures';
 import { Configuration } from '../configuration.js';
+import { getObjectType } from '../debug.js';
 import { Request } from '../request.js';
 import { serviceLocator } from '../service_locator.js';
 import { checkStorageAccess } from './access_checking.js';