fix(oci): polish stream chunk handling and signed body default

- cohere stream terminal consolidation now emits content=None instead of ""
- drop redundant index truthiness check (None is already replaced with 0)
- accept both "TOOL_CALL" and "TOOL_CALLS" finish reasons in cohere
- signed_json_body defaults to None and uses explicit None check, so an
  explicitly empty bytes body wouldn't be silently re-serialized

Co-authored-by: Yassin Kortam <yassin@berri.ai>

litellm/llms/oci/chat/cohere.py

@@ -214,7 +214,7 @@ def handle_cohere_response(
         finish_reason = "stop"
     elif oci_finish_reason == "MAX_TOKENS":
         finish_reason = "length"
-    elif oci_finish_reason == "TOOL_CALL":
+    elif oci_finish_reason in ("TOOL_CALL", "TOOL_CALLS"):
         finish_reason = "tool_calls"
     elif oci_finish_reason is not None:
         # OCI Cohere can emit error/cancel finish reasons (e.g. ``ERROR``,
@@ -291,7 +291,9 @@ def handle_cohere_stream_chunk(dict_chunk: dict) -> ModelResponseStream:
     is_terminal_consolidation = (
         typed_chunk.chatHistory is not None and typed_chunk.finishReason is not None
     )
-    text = "" if is_terminal_consolidation else (typed_chunk.text or "")
+    text: Optional[str] = (
+        None if is_terminal_consolidation else (typed_chunk.text or "")
+    )
 
     # Tool calls on the terminal consolidation chunk (whether from
     # `typed_chunk.toolCalls` or from `chatHistory`) restate what was already
@@ -326,7 +328,7 @@ def handle_cohere_stream_chunk(dict_chunk: dict) -> ModelResponseStream:
         finish_reason = "stop"
     elif finish_reason == "MAX_TOKENS":
         finish_reason = "length"
-    elif finish_reason == "TOOL_CALL":
+    elif finish_reason in ("TOOL_CALL", "TOOL_CALLS"):
         finish_reason = "tool_calls"
     elif finish_reason is not None:
         # OCI Cohere can emit error/cancel finish reasons (e.g. ``ERROR``,
@@ -338,7 +340,7 @@ def handle_cohere_stream_chunk(dict_chunk: dict) -> ModelResponseStream:
     return ModelResponseStream(
         choices=[
             StreamingChoices(
-                index=typed_chunk.index if typed_chunk.index else 0,
+                index=typed_chunk.index,
                 delta=Delta(
                     content=text,
                     tool_calls=tool_calls,

litellm/llms/oci/chat/generic.py

@@ -432,7 +432,7 @@ def handle_generic_stream_chunk(dict_chunk: dict) -> ModelResponseStream:
     return ModelResponseStream(
         choices=[
             StreamingChoices(
-                index=typed_chunk.index if typed_chunk.index else 0,
+                index=typed_chunk.index,
                 delta=Delta(
                     content=text,
                     tool_calls=(

litellm/llms/oci/chat/transformation.py

@@ -565,7 +565,7 @@ class OCIChatConfig(BaseConfig):
         messages: list,
         client: Optional[Union[HTTPHandler, AsyncHTTPHandler]] = None,
         json_mode: Optional[bool] = None,
-        signed_json_body: bytes = b"",
+        signed_json_body: Optional[bytes] = None,
     ) -> "OCIStreamWrapper":
         if "stream" in data:
             del data["stream"]
@@ -576,7 +576,11 @@ class OCIChatConfig(BaseConfig):
             response = client.post(
                 api_base,
                 headers=headers,
-                data=signed_json_body or json.dumps(data),
+                data=(
+                    signed_json_body
+                    if signed_json_body is not None
+                    else json.dumps(data)
+                ),
                 stream=True,
                 logging_obj=logging_obj,
                 timeout=STREAMING_TIMEOUT,
@@ -606,7 +610,7 @@ class OCIChatConfig(BaseConfig):
         messages: list,
         client: Optional[Union[HTTPHandler, AsyncHTTPHandler]] = None,
         json_mode: Optional[bool] = None,
-        signed_json_body: bytes = b"",
+        signed_json_body: Optional[bytes] = None,
     ) -> "OCIStreamWrapper":
         if "stream" in data:
             del data["stream"]
@@ -617,7 +621,11 @@ class OCIChatConfig(BaseConfig):
             response = await client.post(
                 api_base,
                 headers=headers,
-                data=signed_json_body or json.dumps(data),
+                data=(
+                    signed_json_body
+                    if signed_json_body is not None
+                    else json.dumps(data)
+                ),
                 stream=True,
                 logging_obj=logging_obj,
                 timeout=STREAMING_TIMEOUT,

tests/test_litellm/llms/oci/test_oci_coverage_boost.py

@@ -614,7 +614,7 @@ def test_handle_cohere_stream_chunk_complete():
     }
     result = handle_cohere_stream_chunk(chunk)
     assert result.choices[0].finish_reason == "stop"
-    assert result.choices[0].delta.content == ""
+    assert result.choices[0].delta.content is None
 
 
 def test_handle_cohere_stream_chunk_max_tokens():
@@ -626,7 +626,7 @@ def test_handle_cohere_stream_chunk_max_tokens():
     }
     result = handle_cohere_stream_chunk(chunk)
     assert result.choices[0].finish_reason == "length"
-    assert result.choices[0].delta.content == ""
+    assert result.choices[0].delta.content is None
 
 
 def test_handle_cohere_stream_chunk_tool_call():
@@ -638,7 +638,7 @@ def test_handle_cohere_stream_chunk_tool_call():
     }
     result = handle_cohere_stream_chunk(chunk)
     assert result.choices[0].finish_reason == "tool_calls"
-    assert result.choices[0].delta.content == ""
+    assert result.choices[0].delta.content is None
 
 
 def test_handle_cohere_stream_chunk_terminal_drops_full_response_text():
@@ -658,7 +658,7 @@ def test_handle_cohere_stream_chunk_terminal_drops_full_response_text():
         ],
     }
     result = handle_cohere_stream_chunk(chunk)
-    assert result.choices[0].delta.content == ""
+    assert result.choices[0].delta.content is None
 
 
 def test_handle_cohere_stream_chunk_incremental_passes_text_through():