Audio/audio_cli.py

from __future__ import annotations

import argparse
import json
from pathlib import Path


def build_parser() -> argparse.ArgumentParser:
    parser = argparse.ArgumentParser(description="Phill Swarm Audio CLI")
    parser.add_argument("--root-dir", type=str, default=str(Path(__file__).resolve().parent))
    subparsers = parser.add_subparsers(dest="command", required=True)

    subparsers.add_parser("indicators", help="Show Audio runtime and terminal indicators")

    synth = subparsers.add_parser("synthesize", help="Generate speech to a wav file")
    synth.add_argument("--text", required=True)
    synth.add_argument("--output", required=True)
    synth.add_argument("--language", default=None)
    synth.add_argument("--ref-audio", default=None)
    synth.add_argument("--ref-text", default=None)
    synth.add_argument("--instruct", default=None)
    synth.add_argument("--duration", type=float, default=None)
    synth.add_argument("--speed", type=float, default=None)
    synth.add_argument("--num-step", type=int, default=8)
    synth.add_argument("--guidance-scale", type=float, default=1.5)
    synth.add_argument("--denoise", action="store_true")

    speak = subparsers.add_parser("speak", help="Generate speech and try native OS playback")
    speak.add_argument("--text", required=True)
    speak.add_argument("--output", required=True)
    speak.add_argument("--language", default=None)
    speak.add_argument("--ref-audio", default=None)
    speak.add_argument("--ref-text", default=None)
    speak.add_argument("--instruct", default=None)
    speak.add_argument("--duration", type=float, default=None)
    speak.add_argument("--speed", type=float, default=None)
    speak.add_argument("--num-step", type=int, default=8)
    speak.add_argument("--guidance-scale", type=float, default=1.5)
    speak.add_argument("--denoise", action="store_true")
    speak.add_argument("--non-blocking", action="store_true")

    transcribe = subparsers.add_parser("transcribe", help="Transcribe an audio file")
    transcribe.add_argument("--input", required=True)
    return parser


def main() -> None:
    args = build_parser().parse_args()

    if args.command == "synthesize":
        from Audio.Pipeline import load_audio_pipeline

        pipeline = load_audio_pipeline(args.root_dir)
        output = pipeline.synthesize_to_file(
            text=args.text,
            output_path=args.output,
            language=args.language,
            ref_audio=args.ref_audio,
            ref_text=args.ref_text,
            instruct=args.instruct,
            duration=args.duration,
            speed=args.speed,
            num_step=args.num_step,
            guidance_scale=args.guidance_scale,
            denoise=args.denoise,
        )
        print(output)
        return

    if args.command == "speak":
        from Audio.Pipeline import load_audio_pipeline

        pipeline = load_audio_pipeline(args.root_dir)
        output, played = pipeline.speak(
            text=args.text,
            output_path=args.output,
            language=args.language,
            ref_audio=args.ref_audio,
            ref_text=args.ref_text,
            instruct=args.instruct,
            duration=args.duration,
            speed=args.speed,
            num_step=args.num_step,
            guidance_scale=args.guidance_scale,
            denoise=args.denoise,
            block=not args.non_blocking,
        )
        print({"output": str(output), "played": played})
        return

    if args.command == "transcribe":
        from Audio.Pipeline import load_audio_pipeline

        pipeline = load_audio_pipeline(args.root_dir)
        print(pipeline.transcribe_file(args.input))
        return

    if args.command == "indicators":
        from Audio.indicators import build_audio_indicator_payload

        print(json.dumps(build_audio_indicator_payload(args.root_dir), indent=2))


if __name__ == "__main__":
    main()