main.py

import os
from pathlib import Path
from typing import Optional

import typer
from dotenv import load_dotenv
from gtts import gTTS

load_dotenv()

app = typer.Typer(help="Language flashcard generator with TTS and AI enhancement.")


@app.command()
def google(
    text: str = typer.Argument(help="Text to convert to speech."),
    output: Path = typer.Option("output.mp3", "--output", "-o", help="Output file path."),
    lang: str = typer.Option("en", "--lang", "-l", help="Language code (e.g. en, es, fr, ja)."),
    slow: bool = typer.Option(False, "--slow", "-s", help="Speak slowly."),
) -> None:
    """Convert text to speech using Google TTS."""
    tts = gTTS(text=text, lang=lang, slow=slow)
    tts.save(str(output))
    typer.echo(f"Saved audio to {output}")


@app.command()
def azure(
    text: str = typer.Argument(help="Text to convert to speech."),
    output: Path = typer.Option("output.mp3", "--output", "-o", help="Output file path."),
    voice: Optional[str] = typer.Option(None, "--voice", "-v", help="Azure neural voice name (e.g. zu-ZA-ThandoNeural)."),
    lang: str = typer.Option("en-US", "--lang", "-l", help="Language code (e.g. en-US, zu-ZA, ja-JP)."),
    rate: str = typer.Option("medium", "--rate", "-r", help="Speech rate: x-slow, slow, medium, fast, x-fast, or percentage (e.g. +50%, -30%)."),
) -> None:
    """Convert text to speech using Azure TTS."""
    import azure.cognitiveservices.speech as speechsdk
    from azure.identity import DefaultAzureCredential

    endpoint = os.environ.get("SPEECH_ENDPOINT")
    resource_id = os.environ.get("SPEECH_RESOURCE_ID")
    if not endpoint or not resource_id:
        typer.echo("Error: SPEECH_ENDPOINT and SPEECH_RESOURCE_ID environment variables are required.", err=True)
        raise typer.Exit(1)

    speech_config = speechsdk.SpeechConfig(endpoint=endpoint)

    # Authenticate via az cli / DefaultAzureCredential
    aad_token = (
        DefaultAzureCredential()
        .get_token("https://cognitiveservices.azure.com/.default")
        .token
    )
    speech_config.authorization_token = f"aad#{resource_id}#{aad_token}"

    if voice:
        speech_config.speech_synthesis_voice_name = voice
    else:
        speech_config.speech_synthesis_language = lang

    speech_config.set_speech_synthesis_output_format(
        speechsdk.SpeechSynthesisOutputFormat.Audio16Khz32KBitRateMonoMp3
    )

    audio_config = speechsdk.audio.AudioOutputConfig(filename=str(output))
    synthesizer = speechsdk.SpeechSynthesizer(speech_config=speech_config, audio_config=audio_config)

    voice_name = voice or speech_config.speech_synthesis_voice_name
    ssml = (
        f'<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xml:lang="{lang}">'
        f'<voice name="{voice_name}">'
        f'<prosody rate="{rate}">{text}</prosody>'
        f'</voice></speak>'
    )
    result = synthesizer.speak_ssml_async(ssml).get()

    if result.reason == speechsdk.ResultReason.SynthesizingAudioCompleted:
        typer.echo(f"Saved audio to {output}")
    else:
        details = result.cancellation_details
        typer.echo(f"Synthesis failed: {details.reason} — {details.error_details}", err=True)
        raise typer.Exit(1)


@app.command()
def build_deck(
    config: Path = typer.Argument(help="Path to YAML config file (e.g. configs/zulu_common.yaml)."),
    size: Optional[int] = typer.Option(None, "--size", "-s", help="Limit to N cards."),
    render_only: bool = typer.Option(False, "--render-only", help="Re-render .mochi from cached cards only."),
    start: int = typer.Option(0, "--start", help="Start from card index (for resume)."),
) -> None:
    """Build a Mochi flashcard deck from frequency lists using Claude AI and Azure TTS."""
    import yaml

    from enhancer import FlashcardEnhancer
    from models import EnhancementConfig

    if not config.exists():
        typer.echo(f"Config file not found: {config}", err=True)
        raise typer.Exit(1)

    with open(config, "r") as f:
        config_data = yaml.safe_load(f)

    enhancement_config = EnhancementConfig(**config_data)
    enhancer = FlashcardEnhancer(enhancement_config)
    success = enhancer.build(size=size, render_only=render_only, start=start)

    if not success:
        raise typer.Exit(1)


@app.command()
def clean(
    config: Path = typer.Argument(help="Path to YAML config file."),
) -> None:
    """Remove empty (0-byte) audio files left by failed TTS generation."""
    import yaml

    from enhancer import FlashcardEnhancer
    from models import EnhancementConfig

    if not config.exists():
        typer.echo(f"Config file not found: {config}", err=True)
        raise typer.Exit(1)

    with open(config, "r") as f:
        config_data = yaml.safe_load(f)

    enhancement_config = EnhancementConfig(**config_data)
    enhancer = FlashcardEnhancer(enhancement_config)
    enhancer.clean_empty_audio()


if __name__ == "__main__":
    app()