From 611e7f6d173c8f942b10d1cb581e8f8268411377 Mon Sep 17 00:00:00 2001
From: Stephen Zhou <38493346+hyoban@users.noreply.github.com>
Date: Thu, 3 Sep 2026 16:49:44 +0800
Subject: [PATCH] test(knowledge-fs): allow canonical crawl duplicates

---
 ...test_knowledge_fs_initial_source_reliability.py | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/api/tests/unit_tests/services/test_knowledge_fs_initial_source_reliability.py b/api/tests/unit_tests/services/test_knowledge_fs_initial_source_reliability.py
index a4a66b465fa..8b9ff2aa5c7 100644
--- a/api/tests/unit_tests/services/test_knowledge_fs_initial_source_reliability.py
+++ b/api/tests/unit_tests/services/test_knowledge_fs_initial_source_reliability.py
@@ -28,11 +28,15 @@ def test_initial_website_selection_populates_stable_canonical_urls() -> None:
     ]
 
 
-def test_initial_website_selection_rejects_canonical_duplicates() -> None:
-    with pytest.raises(ValidationError, match="canonical URLs must be unique"):
-        KnowledgeFSInitialWebsiteSourcePayload.model_validate(
-            _payload(urls=["https://docs.dify.ai/a/", "https://DOCS.dify.ai:443/a#section"])
-        )
+def test_initial_website_selection_preserves_canonical_duplicates() -> None:
+    payload = KnowledgeFSInitialWebsiteSourcePayload.model_validate(
+        _payload(urls=["https://docs.dify.ai/a/", "https://DOCS.dify.ai:443/a#section"])
+    )
+
+    assert [item.canonical_url for item in payload.selection] == [
+        "https://docs.dify.ai/a",
+        "https://docs.dify.ai/a",
+    ]
 
 
 def test_initial_website_selection_requires_crawl_limit_to_cover_selection() -> None: