Motivation

Llama3.3 frequently fails to generate the <|python_tag|> token at higher temperatures, so additional checks are needed to properly detect tool calls.

Modifications

This commit refactors the tool call detection logic to allow each tool call parser to handle its own detection logic:

1. Added a has_tool_call() method to BaseFormatDetector class
2. Implemented has_tool_call() in Llama32Detector to check for both <|python_tag|> and JSON objects starting with "{"
3. Implemented has_tool_call() in other detector classes (Qwen25Detector, MistralDetector)
4. Updated FunctionCallParser to expose has_tool_call() and check all detectors
5. Modified adapter.py to use the parser's has_tool_call() method instead of hardcoding detection logic

Checklist

• Format your code according to the Code Formatting with Pre-Commit.
• Add unit tests as outlined in the Running Unit Tests.
• Update documentation / docstrings / example tutorials as needed, according to Writing Documentation.
• Provide throughput / latency benchmark results and accuracy evaluation results as needed, according to Benchmark and Profiling and Accuracy Results.
• For reviewers: If you haven't made any contributions to this PR and are only assisting with merging the main branch, please remove yourself as a co-author when merging the PR.
• Please feel free to join our Slack channel at https://slack.sglang.ai to discuss your PR.