fix(ai_model): support reasoning field for Ollama/LMStudio thinking models

Add compatibility for the `reasoning` field in addition to `reasoning_content`.
Ollama/LMStudio follows GPT-OSS official recommendation to use `reasoning`
field for thinking content, while DeepSeek uses `reasoning_content`.

Priority: reasoning_content > reasoning (fallback)

Author: WainWong

backend/apps/ai_model/openai/llm.py

@@ -27,8 +27,12 @@ def _convert_delta_to_message_chunk(
     role = cast(str, _dict.get("role"))
     content = cast(str, _dict.get("content") or "")
     additional_kwargs: dict = {}
-    if 'reasoning_content' in _dict:
-        additional_kwargs['reasoning_content'] = _dict.get('reasoning_content')
+    # 兼容 reasoning_content (DeepSeek等) 和 reasoning (Ollama/LMStudio GPT-OSS) 两种字段
+    reasoning_content = _dict.get('reasoning_content')
+    if not reasoning_content:
+        reasoning_content = _dict.get('reasoning')
+    if reasoning_content:
+        additional_kwargs['reasoning_content'] = reasoning_content
     if _dict.get("function_call"):
         function_call = dict(_dict["function_call"])
         if "name" in function_call and function_call["name"] is None: