# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM-Omni project

"""Online serving benchmark subcommand for vLLM-Omni.

``OmniBenchmarkServingSubcommand`` starts from vLLM's serving benchmark
arguments, adds Omni-specific options and adaptations, then preprocesses the
parsed namespace and delegates execution to ``vllm_omni.benchmarks.serve``.
This keeps CLI integration separate from the multimodal benchmark runtime.
"""

import argparse

from vllm.benchmarks.serve import add_cli_args

from vllm_omni.entrypoints.cli.benchmark.base import OmniBenchmarkSubcommandBase
from vllm_omni.entrypoints.cli.benchmark.cli_args import (
    add_omni_args,
    extend_omni_choices,
    preprocess_serve_args,
    update_omni_help,
)


class OmniBenchmarkServingSubcommand(OmniBenchmarkSubcommandBase):
    """The `serve` subcommand for vllm bench."""

    name = "serve"
    help = "Benchmark online serving. Supports Daily-Omni, OmniInteract, and Seed-TTS datasets."

    @classmethod
    def add_cli_args(cls, parser: argparse.ArgumentParser) -> None:
        add_cli_args(parser)
        add_omni_args(parser)
        extend_omni_choices(parser)
        update_omni_help(parser)

    @staticmethod
    def cmd(args: argparse.Namespace) -> None:
        # Deferred: the benchmark runtime imports the duplex client library
        # (vllm_omni.clients), which must stay out of the CLI import graph —
        # this module is imported by `vllm-omni serve` as well
        # (tests/engine/test_duplex_import_boundary.py enforces the boundary).
        from vllm_omni.benchmarks.serve import main

        preprocess_serve_args(args)
        main(args)