Omit paths and query params when formatting clean URLs for display (#728)

Author: ThomasWCode

web/src/lib/utils/__tests__/formatCleanUrl.test.ts

@@ -15,17 +15,15 @@ describe("formatCleanUrl", () => {
     expect(formatCleanUrl("https://namesake.fyi/")).toBe("namesake.fyi");
   });
 
-  it("removes only a single trailing slash at the end", () => {
-    expect(formatCleanUrl("https://example.com/path/")).toBe(
-      "example.com/path",
-    );
-  });
-
   it("leaves host-only strings unchanged", () => {
     expect(formatCleanUrl("example.com")).toBe("example.com");
   });
 
-  it("returns empty string for empty input", () => {
-    expect(formatCleanUrl("")).toBe("");
+  it("omits the path, query, and fragment", () => {
+    expect(
+      formatCleanUrl(
+        "https://law.uic.edu/experiential-education/clinics/pro-bono/projects/?source=directory#services",
+      ),
+    ).toBe("law.uic.edu");
   });
 });

web/src/lib/utils/formatCleanUrl.ts

@@ -1,10 +1,13 @@
 /**
- * Given a URL, return the URL without protocol, "www", or trailing slash.
+ * Given a URL or hostname, return its hostname without "www".
  *
  * @example
  * formatCleanUrl("https://www.masstpc.org/")
  * // "masstpc.org"
  */
 export function formatCleanUrl(url: string): string {
-  return url.replace(/^https?:\/\/(www\.)?/, "").replace(/\/$/, "");
+  // Ensure handling of links that are already hostnames without typeerror
+  const absoluteUrl = url.includes("://") ? url : `https://${url}`;
+
+  return new URL(absoluteUrl).hostname.replace(/^www\./, "");
 }