@adobe/spacecat-shared-utils 1.44.3 → 1.46.0

package/CHANGELOG.md

@@ -1,3 +1,17 @@
+# [@adobe/spacecat-shared-utils-v1.46.0](https://github.com/adobe/spacecat-shared/compare/@adobe/spacecat-shared-utils-v1.45.0...@adobe/spacecat-shared-utils-v1.46.0) (2025-08-14)
+
+
+### Features
+
+* return null instead of throwing error ([#914](https://github.com/adobe/spacecat-shared/issues/914)) ([90760c0](https://github.com/adobe/spacecat-shared/commit/90760c040c46eab127c1bfc780a454875475fb64))
+
+# [@adobe/spacecat-shared-utils-v1.45.0](https://github.com/adobe/spacecat-shared/compare/@adobe/spacecat-shared-utils-v1.44.3...@adobe/spacecat-shared-utils-v1.45.0) (2025-08-11)
+
+
+### Features
+
+* Create shared utils method for resolving canonical url ([#895](https://github.com/adobe/spacecat-shared/issues/895)) ([054ebf7](https://github.com/adobe/spacecat-shared/commit/054ebf7c0ef90269eac75558d3050ebee712c725))
+
 # [@adobe/spacecat-shared-utils-v1.44.3](https://github.com/adobe/spacecat-shared/compare/@adobe/spacecat-shared-utils-v1.44.2...@adobe/spacecat-shared-utils-v1.44.3) (2025-08-09)
 
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@adobe/spacecat-shared-utils",
-  "version": "1.44.3",
+  "version": "1.46.0",
   "description": "Shared modules of the Spacecat Services - utils",
   "type": "module",
   "engines": {

package/src/index.js

@@ -61,6 +61,9 @@ export {
   stripTrailingDot,
   stripTrailingSlash,
   stripWWW,
+  resolveCanonicalUrl,
+  getSpacecatRequestHeaders,
+  ensureHttps,
 } from './url-helpers.js';
 
 export { getStoredMetrics, storeMetrics } from './metrics-store.js';

package/src/url-helpers.js

@@ -11,6 +11,7 @@
  */
 
 import { context as h2, h1 } from '@adobe/fetch';
+import { SPACECAT_USER_AGENT } from './tracing-fetch.js';
 
 /* c8 ignore next 3 */
 export const { fetch } = process.env.HELIX_FETCH_FORCE_HTTP1
@@ -105,7 +106,75 @@ async function composeAuditURL(url, userAgent) {
   return stripTrailingSlash(finalUrl);
 }
 
+/**
+ * Ensures the URL is HTTPS.
+ * @param {string} url - The URL to ensure is HTTPS.
+ * @returns {string} The HTTPS URL.
+ */
+function ensureHttps(url) {
+  const urlObj = new URL(url);
+  urlObj.protocol = 'https';
+  return urlObj.toString();
+}
+
+/**
+ * Gets spacecat HTTP headers with appropriate user agent for the request type
+ * @returns {Object} - HTTP headers object
+ */
+function getSpacecatRequestHeaders() {
+  return {
+    Accept: 'text/html,application/xhtml+xml,application/xml,text/css,application/javascript,text/javascript;q=0.9,image/avif,image/webp,*/*;q=0.8',
+    'Accept-Language': 'en-US,en;q=0.5',
+    'Cache-Control': 'no-cache',
+    Pragma: 'no-cache',
+    Referer: 'https://www.adobe.com/',
+    'User-Agent': SPACECAT_USER_AGENT,
+  };
+}
+
+/**
+ * Resolve canonical URL for a given URL string by following redirect chain.
+ * @param {string} urlString - The URL string to normalize.
+ * @param {string} method - HTTP method to use ('HEAD' or 'GET').
+ * @returns {Promise<string|null>} A Promise that resolves to the canonical URL or null if failed.
+ */
+async function resolveCanonicalUrl(urlString, method = 'HEAD') {
+  const headers = getSpacecatRequestHeaders();
+  let resp;
+
+  try {
+    resp = await fetch(urlString, { headers, method });
+
+    if (resp.ok) {
+      return ensureHttps(resp.url);
+    }
+
+    // Handle redirect chains
+    if (urlString !== resp.url) {
+      return resolveCanonicalUrl(resp.url, method);
+    }
+
+    if (method === 'HEAD') {
+      return resolveCanonicalUrl(urlString, 'GET');
+    }
+
+    // If the URL is not found and we've tried both HEAD and GET, return null
+    return null;
+  } catch {
+    // If HEAD failed with network error and we haven't tried GET yet, retry with GET
+    if (method === 'HEAD') {
+      return resolveCanonicalUrl(urlString, 'GET');
+    }
+
+    // For all errors (both HTTP status and network), return null
+    return null;
+  }
+}
+
 export {
+  ensureHttps,
+  getSpacecatRequestHeaders,
+  resolveCanonicalUrl,
   composeBaseURL,
   composeAuditURL,
   prependSchema,