Class AudioReader

src/include/audio/audio_reader.hpp:33–60

};

class AudioReader {
public:
    AudioReader();
    ~AudioReader();

    /// Load audio from a file, decode, resample to target_sample_rate, and convert to mono float32.
    /// Supports any format FFmpeg can demux (mp3, wav, flac, ogg, etc.).
    /// @param filename          path to the audio file
    /// @param out_audio         output struct filled with samples and metadata
    /// @param target_sample_rate desired output sample rate (default 16000)
    /// @return true on success
    bool load_audio(const std::string& filename, audio_data_t& out_audio, int target_sample_rate = 16000, MonoDownmixMode downmix = MonoDownmixMode::NONE);

    /// Load audio from an in-memory buffer (e.g. received over network).
    /// @param data              raw file bytes (mp3, wav, etc.)
    /// @param size              byte count
    /// @param out_audio         output struct filled with samples and metadata
    /// @param target_sample_rate desired output sample rate (default 16000)
    /// @param downmix           how to downmix multi-channel audio to mono
    /// @return true on success
    bool load_audio_from_memory(const uint8_t* data, size_t size, audio_data_t& out_audio, int target_sample_rate = 16000, MonoDownmixMode downmix = MonoDownmixMode::NONE);

    // does a simple clipping and simply discarded the audio beyond the max duration.
    bool clip_audio_length(audio_data_t& audio, double max_duration_second);

private:
    bool decode_audio(struct AVFormatContext* format_ctx, audio_data_t& out_audio, int target_sample_rate, MonoDownmixMode downmix);
};