fix(analytics): publish website referrer reporting

frontend/instrumentation-client.ts

@@ -41,7 +41,7 @@ if (typeof window !== "undefined" && posthogConfig) {
     persistence: "memory",
     disable_persistence: true,
     respect_dnt: true,
-    save_referrer: false,
+    save_referrer: true,
     save_campaign_params: false,
     disable_capture_url_hashes: true,
     disable_scroll_properties: true,

frontend/lib/posthog-privacy.ts

@@ -68,6 +68,19 @@ function getRawPath(properties: Record<string, unknown>): unknown {
   }
 }
 
+function sanitizeReferringDomain(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  if (value === "$direct") return value;
+  if (
+    value.length > 253 ||
+    value.trim() !== value ||
+    !/^(?:[a-z0-9](?:[a-z0-9-]{0,61}[a-z0-9])?\.)+[a-z](?:[a-z0-9-]{0,61}[a-z0-9])?$/iu.test(value)
+  ) {
+    return undefined;
+  }
+  return value.toLowerCase();
+}
+
 function createCommonProperties(properties: Record<string, unknown>) {
   const token = properties.token;
   // PostHog's cookieless server-hash ingestion step computes the anonymous
@@ -113,6 +126,8 @@ function createCommonProperties(properties: Record<string, unknown>) {
   if (properties.distinct_id === "$posthog_cookieless") {
     common.distinct_id = "$posthog_cookieless";
   }
+  const referringDomain = sanitizeReferringDomain(properties.$referring_domain);
+  if (referringDomain !== undefined) common.$referring_domain = referringDomain;
   return common;
 }
 

frontend/test/posthog-source.test.mjs

@@ -34,6 +34,7 @@ test("privacy posture disables persistence, recording, autocapture, and automati
     'cookieless_mode: "always"',
     "advanced_disable_flags: true",
     "disable_persistence: true",
+    "save_referrer: true",
     "capture_performance:",
   ]) {
     assert.match(instrumentation, new RegExp(option.replace(/[.*+?^${}()|[\\]\\]/g, "\\$&")));

frontend/test/posthog.test.ts

@@ -1,4 +1,52 @@
 import assert from "node:assert/strict";
+
+test("referring hostnames survive the envelope without URLs or campaign data", () => {
+  const accepted = ["github.com", "news.ycombinator.com", "$direct", "SEARCH.Example"];
+  const rejected = [
+    undefined,
+    null,
+    42,
+    "",
+    "https://search.example/private?email=me@example.com",
+    "//search.example",
+    "search.example/path",
+    "search.example?x=1",
+    "me@example.com",
+    "search.example:443",
+    "search.example\n",
+    "-bad.example",
+    "bad-.example",
+    `${"a".repeat(64)}.example`,
+  ];
+  for (const event of ["$pageview", "$pageleave", "$web_vitals"]) {
+    for (const value of [...accepted, ...rejected]) {
+      const result = sanitizeEvent({
+        event,
+        properties: {
+          token: "phc_referrer_fixture",
+          $cookieless_mode: true,
+          $process_person_profile: false,
+          $raw_user_agent: "Mozilla/5.0 (Referrer Fixture)",
+          $host: "codeswhat.com",
+          path: "/",
+          $web_vitals_LCP_value: 100,
+          $referring_domain: value,
+          $referrer: "https://search.example/private?email=me@example.com",
+          utm_source: "me@example.com",
+        },
+      });
+      assert.ok(result);
+      assert.equal(
+        result.properties.$referring_domain,
+        accepted.includes(value as string) ? (value as string).toLowerCase() : undefined,
+        `${event}: ${JSON.stringify(value)}`,
+      );
+      assert.equal("$referrer" in result.properties, false);
+      assert.equal("utm_source" in result.properties, false);
+    }
+  }
+});
+
 import { createRequire } from "node:module";
 import { test } from "node:test";
 import {