tangara-fw/src/audio/audio_decoder.cpp

#include "audio_decoder.hpp"

#include <string.h>

#include <cstddef>
#include <cstdint>

#include "freertos/FreeRTOS.h"

#include "esp_heap_caps.h"
#include "freertos/message_buffer.h"
#include "freertos/portmacro.h"

#include "audio_element.hpp"
#include "chunk.hpp"
#include "fatfs_audio_input.hpp"

static const char* kTag = "DEC";

namespace audio {

AudioDecoder::AudioDecoder()
    : IAudioElement(),
      stream_info_({}),
      chunk_buffer_(static_cast<uint8_t*>(
          heap_caps_malloc(kMaxChunkSize, MALLOC_CAP_SPIRAM)))

{}

AudioDecoder::~AudioDecoder() {
  free(chunk_buffer_);
}

auto AudioDecoder::SetInputBuffer(MessageBufferHandle_t* buffer) -> void {
  input_buffer_ = buffer;
}

auto AudioDecoder::SetOutputBuffer(MessageBufferHandle_t* buffer) -> void {
  output_buffer_ = buffer;
}

auto AudioDecoder::ProcessStreamInfo(StreamInfo&& info)
    -> cpp::result<void, StreamError> {
  stream_info_ = info;

  // Reuse the existing codec if we can. This will help with gapless playback,
  // since we can potentially just continue to decode as we were before,
  // without any setup overhead.
  if (current_codec_->CanHandleFile(info.Path().value_or(""))) {
    current_codec_->ResetForNewStream();
    return {};
  }

  auto result = codecs::CreateCodecForFile(info.Path().value());
  if (result.has_value()) {
    current_codec_ = std::move(result.value());
  } else {
    return cpp::fail(UNSUPPORTED_STREAM);
  }

  return {};
}

auto AudioDecoder::ProcessChunk(uint8_t* data, std::size_t length)
    -> cpp::result<size_t, StreamError> {
  if (current_codec_ == nullptr) {
    // Should never happen, but fail explicitly anyway.
    return cpp::fail(UNSUPPORTED_STREAM);
  }

  current_codec_->SetInput(data, length);

  bool has_samples_to_send = false;
  bool needs_more_input = false;
  std::optional<codecs::ICodec::ProcessingError> error = std::nullopt;
  WriteChunksToStream(
      output_buffer_, chunk_buffer_, kMaxChunkSize,
      [&](uint8_t* buf, size_t len) -> std::size_t {
        std::size_t bytes_written = 0;
        // Continue filling up the output buffer so long as we have samples
        // leftover, or are able to synthesize more samples from the input.
        while (has_samples_to_send || !needs_more_input) {
          if (!has_samples_to_send) {
            auto result = current_codec_->ProcessNextFrame();
            has_samples_to_send = true;
            if (result.has_error()) {
              error = result.error();
              // End our output stream immediately if the codec barfed.
              return 0;
            } else {
              needs_more_input = result.value();
            }
          } else {
            auto result = current_codec_->WriteOutputSamples(
                buf + bytes_written, len - bytes_written);
            bytes_written += result.first;
            has_samples_to_send = !result.second;
          }
        }
        return bytes_written;
      },
      // This element doesn't support any kind of out of band commands, so we
      // can just suspend the whole task if the output buffer fills up.
      portMAX_DELAY);

  if (error) {
    ESP_LOGE(kTag, "Codec encountered error %d", error.value());
    return cpp::fail(IO_ERROR);
  }

  return current_codec_->GetInputPosition();
}

auto AudioDecoder::ProcessIdle() -> cpp::result<void, StreamError> {
  // Not used; we delay forever when waiting on IO.
  return {};
}

}  // namespace audio