Report full URL instead of bare origin for detected scripts and iframes

Strip query params and send origin+pathname so the backend can
distinguish resources served from the same domain but different paths
(e.g. gtm.js vs recaptcha/api.js on googletagmanager.com).

Signed-off-by: Émile Ré <emile@getprobo.com>

packages/cookie-banner/src/detectors/third-party-detector.ts

@@ -17,7 +17,7 @@ import { NotFoundError } from "../errors";
 import { fetchJSON } from "../http";
 
 interface DetectedResourceEntry {
-  origin: string;
+  url: string;
   resource_type: "script" | "iframe";
 }
 
@@ -105,20 +105,21 @@ export class ThirdPartyDetector implements Detector {
   private processElement(src: string, resourceType: "script" | "iframe"): void {
     if (EXTENSION_URL_RE.test(src)) return;
 
-    let origin: string;
+    let parsed: URL;
     try {
-      origin = new URL(src).origin;
+      parsed = new URL(src);
     } catch {
       return;
     }
 
-    if (origin === this.pageOrigin || origin === this.proboOrigin) return;
+    if (parsed.origin === this.pageOrigin || parsed.origin === this.proboOrigin) return;
 
-    const reportKey = `${resourceType}:${origin}`;
+    const identifier = parsed.origin + parsed.pathname;
+    const reportKey = `${resourceType}:${identifier}`;
     if (this.reported.has(reportKey)) return;
 
     this.reported.add(reportKey);
-    this.pending.set(reportKey, { origin, resource_type: resourceType });
+    this.pending.set(reportKey, { url: identifier, resource_type: resourceType });
     this.scheduleFlush();
   }
 

pkg/cookiebanner/service.go

@@ -30,6 +30,7 @@ import (
 	"go.probo.inc/probo/pkg/coredata"
 	"go.probo.inc/probo/pkg/gid"
 	"go.probo.inc/probo/pkg/page"
+	"go.probo.inc/probo/pkg/uri"
 	"go.probo.inc/probo/pkg/validator"
 )
 
@@ -125,7 +126,7 @@ type (
 	}
 
 	DetectedResourceItem struct {
-		Origin       string
+		URL          uri.URI
 		ResourceType coredata.TrackerType
 	}
 
@@ -1992,7 +1993,7 @@ func (s *Service) ReportDetectedTrackers(
 					now,
 					detectedTrackerInfo{
 						TrackerType: dr.ResourceType,
-						Identifier:  dr.Origin,
+						Identifier:  dr.URL.String(),
 					},
 					&inserted,
 				); err != nil {

pkg/server/api/cookiebanner/v1/handler.go

@@ -31,6 +31,7 @@ import (
 	"go.probo.inc/probo/pkg/gid"
 	"go.probo.inc/probo/pkg/server/api/clientip"
 	"go.probo.inc/probo/pkg/server/jsonutil"
+	"go.probo.inc/probo/pkg/uri"
 )
 
 type Handler struct {
@@ -309,8 +310,8 @@ type detectedStorageEntry struct {
 }
 
 type detectedResourceEntry struct {
-	Origin       string `json:"origin"`
-	ResourceType string `json:"resource_type"`
+	URL          uri.URI `json:"url"`
+	ResourceType string  `json:"resource_type"`
 }
 
 type reportDetectedTrackersBody struct {
@@ -400,11 +401,6 @@ func (h *Handler) handleReportDetectedTrackers(w http.ResponseWriter, r *http.Re
 	}
 
 	for _, res := range body.Resources {
-		origin := strings.TrimSpace(res.Origin)
-		if origin == "" {
-			continue
-		}
-
 		var resourceType coredata.TrackerType
 		switch strings.TrimSpace(res.ResourceType) {
 		case "script":
@@ -418,7 +414,7 @@ func (h *Handler) handleReportDetectedTrackers(w http.ResponseWriter, r *http.Re
 		req.Resources = append(
 			req.Resources,
 			cookiebanner.DetectedResourceItem{
-				Origin:       origin,
+				URL:          res.URL,
 				ResourceType: resourceType,
 			},
 		)