[frontend] enhance tool_calls type check

[reidliu41]

FIX #16678 (link existing issues this PR will resolve)

Based on the description with tool_calls is None, it can reproduce:

client:
python examples/online_serving/openai_chat_completion_client_with_tools.py

    raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'object': 'error', 'message': "'NoneType' object is not iterable", 'type': 'BadRequestError', 'param': None, 'code': 400}


server:
INFO 04-20 07:58:28 [async_llm.py:239] Added request chatcmpl-c45e0e16c6694176a2ca86df9926de39.
ERROR 04-20 07:58:29 [serving_chat.py:200] Error in preprocessing prompt inputs
ERROR 04-20 07:58:29 [serving_chat.py:200] Traceback (most recent call last):
ERROR 04-20 07:58:29 [serving_chat.py:200]   File "/home/user/vllm/vllm/entrypoints/openai/serving_chat.py", line 183, in create_chat_completion
ERROR 04-20 07:58:29 [serving_chat.py:200]     ) = await self._preprocess_chat(
ERROR 04-20 07:58:29 [serving_chat.py:200]         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 04-20 07:58:29 [serving_chat.py:200]   File "/home/user/vllm/vllm/entrypoints/openai/serving_engine.py", line 403, in _preprocess_chat
ERROR 04-20 07:58:29 [serving_chat.py:200]     conversation, mm_data_future = parse_chat_messages_futures(
ERROR 04-20 07:58:29 [serving_chat.py:200]                                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 04-20 07:58:29 [serving_chat.py:200]   File "/home/user/vllm/vllm/entrypoints/chat_utils.py", line 1158, in parse_chat_messages_futures
ERROR 04-20 07:58:29 [serving_chat.py:200]     sub_messages = _parse_chat_message_content(
ERROR 04-20 07:58:29 [serving_chat.py:200]                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 04-20 07:58:29 [serving_chat.py:200]   File "/home/user/vllm/vllm/entrypoints/chat_utils.py", line 1098, in _parse_chat_message_content
ERROR 04-20 07:58:29 [serving_chat.py:200]     result_msg["tool_calls"] = list(parsed_msg["tool_calls"])
ERROR 04-20 07:58:29 [serving_chat.py:200]                                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 04-20 07:58:29 [serving_chat.py:200] TypeError: 'NoneType' object is not iterable

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.

💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels.

Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can either: Add ready label to the PR or enable auto-merge.

🚀

[reidliu41]

@DarkLight1337 could you pls also help to take a look this?

[DarkLight1337]

Can you add a comment explaining that this is just for compatibility purposes? Since it's technically unnecessary if downstream compiles with OpenAI spec