Merge 1e02f5dafd into f3d4605dc7

api/core/rag/extractor/pdf_extractor.py

@@ -115,7 +115,7 @@ class PdfExtractor(BaseExtractor):
         """
         image_content = []
         upload_files = []
-        base_url = dify_config.INTERNAL_FILES_URL or dify_config.FILES_URL
+        base_url = dify_config.FILES_URL
 
         try:
             image_objects = page.get_objects(filter=(pdfium_c.FPDF_PAGEOBJ_IMAGE,))

api/core/rag/extractor/word_extractor.py

@@ -88,7 +88,7 @@ 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
+        base_url = dify_config.FILES_URL
 
         for r_id, rel in doc.part.rels.items():
             if "image" in rel.target_ref:

api/core/tools/signature.py

@@ -25,12 +25,15 @@ def sign_tool_file(tool_file_id: str, extension: str, for_external: bool = True)
     return f"{file_preview_url}?timestamp={timestamp}&nonce={nonce}&sign={encoded_sign}"
 
 
-def sign_upload_file(upload_file_id: str, extension: str) -> str:
+def sign_upload_file(upload_file_id: str, extension: str, for_external: bool = True) -> str:
     """
-    sign file to get a temporary url for plugin access
+    Sign an upload file to get a temporary image preview URL.
+
+    External URLs are the default because uploaded-file previews are returned to
+    user-facing retrieval responses. Internal URLs remain available for callers
+    that need to access files inside the deployment network.
     """
-    # Use internal URL for plugin/tool file access in Docker environments
-    base_url = dify_config.INTERNAL_FILES_URL or dify_config.FILES_URL
+    base_url = dify_config.FILES_URL if for_external else (dify_config.INTERNAL_FILES_URL or dify_config.FILES_URL)
     file_preview_url = f"{base_url}/files/{upload_file_id}/image-preview"
 
     timestamp = str(int(time.time()))

api/models/dataset.py

@@ -965,7 +965,7 @@ class DocumentSegment(Base):
             encoded_sign = base64.urlsafe_b64encode(sign).decode()
 
             params = f"timestamp={timestamp}&nonce={nonce}&sign={encoded_sign}"
-            reference_url = dify_config.CONSOLE_API_URL or ""
+            reference_url = dify_config.FILES_URL or dify_config.CONSOLE_API_URL or ""
             base_url = f"{reference_url}/files/{upload_file_id}/image-preview"
             source_url = f"{base_url}?{params}"
             attachment_list.append(

api/tests/unit_tests/core/tools/test_signature.py

@@ -83,7 +83,7 @@ def test_verify_tool_file_signature_rejects_expired_signature(monkeypatch: pytes
     assert verify_tool_file_signature("tool-file-id", timestamp, nonce, sign) is False
 
 
-def test_sign_upload_file_prefers_internal_url(monkeypatch: pytest.MonkeyPatch) -> None:
+def test_sign_upload_file_for_external_uses_files_url(monkeypatch: pytest.MonkeyPatch) -> None:
     monkeypatch.setattr("core.tools.signature.time.time", lambda: 1700000000)
     monkeypatch.setattr("core.tools.signature.os.urandom", lambda _: b"\x03" * 16)
     monkeypatch.setattr("core.tools.signature.dify_config.SECRET_KEY", "unit-secret")
@@ -94,6 +94,24 @@ def test_sign_upload_file_prefers_internal_url(monkeypatch: pytest.MonkeyPatch)
     parsed = urlparse(url)
     query = parse_qs(parsed.query)
 
+    assert parsed.netloc == "files.example.com"
+    assert parsed.path == "/files/upload-id/image-preview"
+    assert query["timestamp"][0]
+    assert query["nonce"][0]
+    assert query["sign"][0]
+
+
+def test_sign_upload_file_for_internal_prefers_internal_url(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr("core.tools.signature.time.time", lambda: 1700000000)
+    monkeypatch.setattr("core.tools.signature.os.urandom", lambda _: b"\x08" * 16)
+    monkeypatch.setattr("core.tools.signature.dify_config.SECRET_KEY", "unit-secret")
+    monkeypatch.setattr("core.tools.signature.dify_config.FILES_URL", "https://files.example.com")
+    monkeypatch.setattr("core.tools.signature.dify_config.INTERNAL_FILES_URL", "https://internal.example.com")
+
+    url = sign_upload_file("upload-id", ".png", for_external=False)
+    parsed = urlparse(url)
+    query = parse_qs(parsed.query)
+
     assert parsed.netloc == "internal.example.com"
     assert parsed.path == "/files/upload-id/image-preview"
     assert query["timestamp"][0]