feat: first use INTERNAL_FILES_URL first, then FILES_URL (#29962)

api/core/rag/extractor/word_extractor.py

@@ -83,6 +83,7 @@ class WordExtractor(BaseExtractor):
     def _extract_images_from_docx(self, doc):
         image_count = 0
         image_map = {}
+        base_url = dify_config.INTERNAL_FILES_URL or dify_config.FILES_URL
 
         for r_id, rel in doc.part.rels.items():
             if "image" in rel.target_ref:
@@ -121,8 +122,7 @@ class WordExtractor(BaseExtractor):
                             used_at=naive_utc_now(),
                         )
                         db.session.add(upload_file)
-                        # Use r_id as key for external images since target_part is undefined
+                        image_map[r_id] = f"![image]({base_url}/files/{upload_file.id}/file-preview)"
-                        image_map[r_id] = f"![image]({dify_config.FILES_URL}/files/{upload_file.id}/file-preview)"
                 else:
                     image_ext = rel.target_ref.split(".")[-1]
                     if image_ext is None:
@@ -150,10 +150,7 @@ class WordExtractor(BaseExtractor):
                         used_at=naive_utc_now(),
                     )
                     db.session.add(upload_file)
-                    # Use target_part as key for internal images
+                    image_map[rel.target_part] = f"![image]({base_url}/files/{upload_file.id}/file-preview)"
-                    image_map[rel.target_part] = (
-                        f"![image]({dify_config.FILES_URL}/files/{upload_file.id}/file-preview)"
-                    )
         db.session.commit()
         return image_map
 

api/tests/unit_tests/core/rag/extractor/test_word_extractor.py

@@ -132,3 +132,36 @@ def test_extract_images_from_docx(monkeypatch):
     # DB interactions should be recorded
     assert len(db_stub.session.added) == 2
     assert db_stub.session.committed is True
+
+
+def test_extract_images_from_docx_uses_internal_files_url():
+    """Test that INTERNAL_FILES_URL takes precedence over FILES_URL for plugin access."""
+    # Test the URL generation logic directly
+    from configs import dify_config
+
+    # Mock the configuration values
+    original_files_url = getattr(dify_config, "FILES_URL", None)
+    original_internal_files_url = getattr(dify_config, "INTERNAL_FILES_URL", None)
+
+    try:
+        # Set both URLs - INTERNAL should take precedence
+        dify_config.FILES_URL = "http://external.example.com"
+        dify_config.INTERNAL_FILES_URL = "http://internal.docker:5001"
+
+        # Test the URL generation logic (same as in word_extractor.py)
+        upload_file_id = "test_file_id"
+
+        # This is the pattern we fixed in the word extractor
+        base_url = dify_config.INTERNAL_FILES_URL or dify_config.FILES_URL
+        generated_url = f"{base_url}/files/{upload_file_id}/file-preview"
+
+        # Verify that INTERNAL_FILES_URL is used instead of FILES_URL
+        assert "http://internal.docker:5001" in generated_url, f"Expected internal URL, got: {generated_url}"
+        assert "http://external.example.com" not in generated_url, f"Should not use external URL, got: {generated_url}"
+
+    finally:
+        # Restore original values
+        if original_files_url is not None:
+            dify_config.FILES_URL = original_files_url
+        if original_internal_files_url is not None:
+            dify_config.INTERNAL_FILES_URL = original_internal_files_url