Public declaration syntax from forge-rs/crates/forge-media/src/speech.rs
Original source SHA-256: 70552b7d5e97993f9f1ecf278efe005f90936e1402dd4c207c3c5946caeae6f7
Function bodies and constant values are omitted. This is not the complete implementation.

Source line 48
#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash, Default, Serialize, Deserialize)]
#[serde(rename_all = "lowercase")]
pub enum AudioFormat {
    /// MP3 — widely supported lossy audio format.
    #[default]
    Mp3,
    /// WAV — uncompressed PCM audio.
    Wav,
    /// OGG — open container format, typically with Vorbis or Opus codec.
    Ogg,
    /// FLAC — lossless audio compression.
    Flac,
}

Source line 77
pub fn mime_type(&self) -> &'static str;

Source line 98
pub fn extension(&self) -> &'static str;

Source line 133
#[derive(Debug, Clone, Default, Serialize, Deserialize)]
pub struct SpeechOptions {
/// The voice identifier to use (e.g., "alloy", "echo", "nova").

/// `None` means the provider's default voice.

pub voice: Option<String>,
/// The playback speed multiplier (e.g., 0.5 for half speed, 2.0 for double speed).

/// `None` means the provider's default speed (typically 1.0).

pub speed: Option<f64>,
/// Output audio format.

pub format: AudioFormat
}

Source line 163
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SpeechResult {
/// The raw audio bytes in the format specified by `mime_type`.

pub audio: Vec<u8>,
/// The MIME type of the audio data (e.g., "audio/mpeg").

pub mime_type: String,
/// The duration of the generated audio in seconds, if available.

/// `None` if the provider does not report duration.

pub duration_seconds: Option<f64>
}

Source line 213
#[async_trait]
pub trait SpeechProvider: Send + Sync {
    /// Returns the model identifier (e.g., "tts-1", "tts-1-hd").
    fn model_id(&self) -> &str;

    /// Returns the provider name (e.g., "openai", "elevenlabs").
    fn provider_name(&self) -> &str;

    /// Synthesizes speech from text.
    ///
    /// # Arguments
    ///
    /// * `text` - The text to convert to speech.
    /// * `options` - Configuration for voice, speed, and output format.
    ///
    /// # Returns
    ///
    /// A [`SpeechResult`] containing the generated audio bytes and metadata.
    ///
    /// # Errors
    ///
    /// * [`ForgeMediaError::SpeechFailed`] -- if the provider returns an error.
    /// * [`ForgeMediaError::UnsupportedFormat`] -- if the requested format is not supported.
    async fn speak(
        &self,
        text: &str,
        options: &SpeechOptions,
    ) -> Result<SpeechResult, ForgeMediaError>;
}

Source line 283
pub async fn speak(
    provider: &dyn SpeechProvider,
    text: &str,
    options: &SpeechOptions,
) -> Result<SpeechResult, ForgeMediaError>;