fix: preserve non-image file MIME type in LiteLlm content conversion

When converting an inlined artifact Part to a LiteLlm/OpenAI file block,
the MIME type of non-image files was dropped and sent as None, so OpenAI
rejected PDFs loaded via load_artifacts (while Gemini and Claude worked).
Propagate the part's mime_type into the file content block.

Close #4174

Co-authored-by: George Weale <gweale@google.com>
PiperOrigin-RevId: 949061584

src/google/adk/models/lite_llm.py

@@ -1329,7 +1329,7 @@ async def _get_content(
           )
           content_objects.append({
               "type": "file",
-              "file": {"file_id": file_response.id},
+              "file": {"file_id": file_response.id, "format": mime_type},
           })
         else:
           content_objects.append({

tests/unittests/models/test_litellm.py

@@ -5104,6 +5104,7 @@ async def test_get_content_pdf_openai_uses_file_id(mocker):
 
   assert content[0]["type"] == "file"
   assert content[0]["file"]["file_id"] == "file-abc123"
+  assert content[0]["file"]["format"] == "application/pdf"
   assert "file_data" not in content[0]["file"]
 
   mock_acreate_file.assert_called_once_with(
@@ -5144,6 +5145,7 @@ async def test_get_content_pdf_azure_uses_file_id(mocker):
 
   assert content[0]["type"] == "file"
   assert content[0]["file"]["file_id"] == "file-xyz789"
+  assert content[0]["file"]["format"] == "application/pdf"
 
   mock_acreate_file.assert_called_once_with(
       file=b"test_pdf_data",
@@ -5189,6 +5191,7 @@ async def test_get_completion_inputs_openai_file_upload(mocker):
   assert content[0]["text"] == "Analyze this PDF"
   assert content[1]["type"] == "file"
   assert content[1]["file"]["file_id"] == "file-uploaded123"
+  assert content[1]["file"]["format"] == "application/pdf"
 
   mock_acreate_file.assert_called_once()