[Style] Collapse kv cache coordinator MTP guard to one line

ruff format collapses the multi-line condition in the multi-module MTP
guard since it fits within the 120-char line length. Fixes the pre-commit
ruff-format hook and the ci-gate it gates.

Signed-off-by: liaoqidan <1107297340@qq.com>

Author: LQDLove

vllm_ascend/patch/platform/patch_kv_cache_coordinator.py

@@ -138,11 +138,7 @@ def __init__(
         # Multi-module MTP with prefix caching requires the scheduler block to
         # hold the whole prefill lookahead, mirroring the upstream coordinator
         # contract from vllm-project/vllm#50062.
-        if (
-            enable_caching
-            and self.eagle_group_ids
-            and scheduler_block_size < num_prefill_lookahead
-        ):
+        if enable_caching and self.eagle_group_ids and scheduler_block_size < num_prefill_lookahead:
             raise ValueError(
                 "Multi-module MTP with prefix caching requires scheduler_block_size"
                 f" (={scheduler_block_size}) >= num_speculative_tokens"