[v0][structured output] Support reasoning output (#12955)

Signed-off-by: Ce Gao <cegao@tensorchord.ai>

vllm/config.py

@@ -2715,6 +2715,8 @@ class DecodingConfig:
     # 'outlines' / 'lm-format-enforcer' / 'xgrammar'
     guided_decoding_backend: str = 'xgrammar'
 
+    reasoning_backend: Optional[str] = None
+
     def compute_hash(self) -> str:
         """
         WARNING: Whenever a new field is added to this config,