diff --git a/astrbot/core/provider/sources/openai_source.py b/astrbot/core/provider/sources/openai_source.py index 67971a2a93..eed4f533d3 100644 --- a/astrbot/core/provider/sources/openai_source.py +++ b/astrbot/core/provider/sources/openai_source.py @@ -69,6 +69,38 @@ def _safe_json_dump(value: Any) -> str | None: except Exception: return None + @staticmethod + def _is_empty_assistant_content(content: Any) -> bool: + if isinstance(content, str): + return content.strip() == "" + return content is None or content == [] + + @classmethod + def _clean_assistant_messages_for_provider(cls, payloads: dict) -> None: + messages = payloads.get("messages") + if not isinstance(messages, list): + return + + cleaned_messages = [] + for idx, msg in enumerate(messages): + if msg.get("role") == "assistant": + content = msg.get("content") + tool_calls = msg.get("tool_calls") + + if not tool_calls and cls._is_empty_assistant_content(content): + logger.debug( + f"Filtered empty assistant message at index {idx} " + "(without tool calls)" + ) + continue + + if tool_calls and cls._is_empty_assistant_content(content): + msg["content"] = None + + cleaned_messages.append(msg) + + payloads["messages"] = cleaned_messages + def _get_image_moderation_error_patterns(self) -> list[str]: """Return configured moderation patterns (case-insensitive substring match, not regex).""" configured = self.provider_config.get("image_moderation_error_patterns", []) @@ -548,26 +580,7 @@ async def _query(self, payloads: dict, tools: ToolSet | None) -> LLMResponse: model = payloads.get("model", "").lower() - if "messages" in payloads and isinstance(payloads["messages"], list): - cleaned_messages = [] - for idx, msg in enumerate(payloads["messages"]): - # 过滤空的 assistant 消息，防止严格 API（如 Moonshot）返回 400 错误 - if msg.get("role") == "assistant": - content = msg.get("content") - tool_calls = msg.get("tool_calls") - - # 情况1: 空/null content 且无 tool_calls -> 过滤掉 - if not tool_calls and (content == "" or content is None): - logger.warning(f"过滤第 {idx} 条空 assistant 消息 (无工具调用)") - continue - - # 情况2: 空 content 但有 tool_calls -> 设为 None (符合 OpenAI 规范) - if content == "" and tool_calls: - msg["content"] = None - - cleaned_messages.append(msg) - - payloads["messages"] = cleaned_messages + self._clean_assistant_messages_for_provider(payloads) completion = await self.client.chat.completions.create( **payloads, @@ -618,6 +631,7 @@ async def _query_stream( for key in to_del: del payloads[key] self._apply_provider_specific_extra_body_overrides(extra_body) + self._clean_assistant_messages_for_provider(payloads) stream = await self.client.chat.completions.create( **payloads, diff --git a/tests/test_openai_source.py b/tests/test_openai_source.py index 83e18137c4..cdc6922001 100644 --- a/tests/test_openai_source.py +++ b/tests/test_openai_source.py @@ -1263,6 +1263,98 @@ async def fake_create(**kwargs): await provider.terminate() +@pytest.mark.asyncio +async def test_query_stream_filters_empty_assistant_messages(monkeypatch): + provider = _make_provider() + try: + captured_kwargs = {} + chunks = [ + ChatCompletionChunk.model_validate( + { + "id": "chatcmpl-stream", + "object": "chat.completion.chunk", + "created": 0, + "model": "gpt-4o-mini", + "choices": [ + { + "index": 0, + "delta": { + "role": "assistant", + "content": "ok", + }, + "finish_reason": None, + } + ], + } + ), + ChatCompletionChunk.model_validate( + { + "id": "chatcmpl-stream", + "object": "chat.completion.chunk", + "created": 0, + "model": "gpt-4o-mini", + "choices": [ + { + "index": 0, + "delta": {}, + "finish_reason": "stop", + } + ], + } + ), + ] + + async def fake_stream(): + for chunk in chunks: + yield chunk + + async def fake_create(**kwargs): + captured_kwargs.update(kwargs) + return fake_stream() + + monkeypatch.setattr(provider.client.chat.completions, "create", fake_create) + + tool_calls = [ + { + "id": "call-1", + "type": "function", + "function": { + "name": "get_weather", + "arguments": "{}", + }, + } + ] + payloads = { + "model": "deepseek-reasoner", + "messages": [ + {"role": "user", "content": "hello"}, + {"role": "assistant", "content": ""}, + {"role": "assistant", "content": None}, + {"role": "assistant", "content": []}, + {"role": "assistant", "content": " "}, + {"role": "assistant", "content": "", "tool_calls": tool_calls}, + {"role": "assistant", "content": " ", "tool_calls": tool_calls}, + {"role": "user", "content": "world"}, + ], + } + + responses = [ + response + async for response in provider._query_stream(payloads=payloads, tools=None) + ] + + messages = captured_kwargs["messages"] + assert responses[-1].completion_text == "ok" + assert messages == [ + {"role": "user", "content": "hello"}, + {"role": "assistant", "content": None, "tool_calls": tool_calls}, + {"role": "assistant", "content": None, "tool_calls": tool_calls}, + {"role": "user", "content": "world"}, + ] + finally: + await provider.terminate() + + @pytest.mark.asyncio async def test_query_filters_empty_assistant_message_without_tool_calls(monkeypatch): """Test that empty assistant messages without tool_calls are filtered out.""" @@ -1303,6 +1395,7 @@ async def fake_create(**kwargs): "messages": [ {"role": "user", "content": "hello"}, {"role": "assistant", "content": ""}, # Should be filtered + {"role": "assistant", "content": " "}, # Should be filtered {"role": "user", "content": "world"}, ], } @@ -1375,6 +1468,62 @@ async def fake_create(**kwargs): await provider.terminate() +@pytest.mark.asyncio +async def test_query_filters_empty_list_assistant_message_without_tool_calls( + monkeypatch, +): + """Test that assistant messages with empty list content are filtered.""" + provider = _make_provider() + try: + captured_kwargs = {} + + async def fake_create(**kwargs): + captured_kwargs.update(kwargs) + return ChatCompletion.model_validate( + { + "id": "chatcmpl-test", + "object": "chat.completion", + "created": 0, + "model": "gpt-4o-mini", + "choices": [ + { + "index": 0, + "message": { + "role": "assistant", + "content": "ok", + }, + "finish_reason": "stop", + } + ], + "usage": { + "prompt_tokens": 1, + "completion_tokens": 1, + "total_tokens": 2, + }, + } + ) + + monkeypatch.setattr(provider.client.chat.completions, "create", fake_create) + + payloads = { + "model": "gpt-4o-mini", + "messages": [ + {"role": "user", "content": "hello"}, + {"role": "assistant", "content": []}, # Should be filtered + {"role": "user", "content": "world"}, + ], + } + + await provider._query(payloads=payloads, tools=None) + + messages = captured_kwargs["messages"] + assert len(messages) == 2 + assert messages[0] == {"role": "user", "content": "hello"} + assert messages[1] == {"role": "user", "content": "world"} + finally: + await provider.terminate() + + @pytest.mark.asyncio async def test_query_converts_empty_content_to_none_with_tool_calls(monkeypatch): """Test that empty content with tool_calls is converted to None (OpenAI spec)."""