video-pc/AvRecorder/capturer/audio/audio_qt_capturer.cpp

#include "audio_qt_capturer.h"
#include "qaudio.h"

#include <QAudioDeviceInfo>
#include <QDebug>

#include <qendian.h>

AudioInfo::AudioInfo(const QAudioFormat& format)
    : m_format(format)
{
    switch (m_format.sampleSize()) {
    case 8:
        switch (m_format.sampleType()) {
        case QAudioFormat::UnSignedInt:
            m_maxAmplitude = 255;
            break;
        case QAudioFormat::SignedInt:
            m_maxAmplitude = 127;
            break;
        default:
            break;
        }
        break;
    case 16:
        switch (m_format.sampleType()) {
        case QAudioFormat::UnSignedInt:
            m_maxAmplitude = 65535;
            break;
        case QAudioFormat::SignedInt:
            m_maxAmplitude = 32767;
            break;
        default:
            break;
        }
        break;

    case 32:
        switch (m_format.sampleType()) {
        case QAudioFormat::UnSignedInt:
            m_maxAmplitude = 0xffffffff;
            break;
        case QAudioFormat::SignedInt:
            m_maxAmplitude = 0x7fffffff;
            break;
        case QAudioFormat::Float:
            m_maxAmplitude = 0x7fffffff; // Kind of
        default:
            break;
        }
        break;

    default:
        break;
    }
}

void AudioInfo::start()
{
    open(QIODevice::WriteOnly);
}

void AudioInfo::stop()
{
    close();
}

qint64 AudioInfo::readData(char* data, qint64 maxlen)
{
    Q_UNUSED(data)
    Q_UNUSED(maxlen)

    return 0;
}

qint64 AudioInfo::writeData(const char* data, qint64 len)
{
    if (m_maxAmplitude) {
        Q_ASSERT(m_format.sampleSize() % 8 == 0);
        const int channelBytes = m_format.sampleSize() / 8;
        const int sampleBytes = m_format.channelCount() * channelBytes;
        Q_ASSERT(len % sampleBytes == 0);
        const int numSamples = len / sampleBytes;

        quint32 maxValue = 0;
        const unsigned char* ptr = reinterpret_cast<const unsigned char*>(data);

        for (int i = 0; i < numSamples; ++i) {
            for (int j = 0; j < m_format.channelCount(); ++j) {
                quint32 value = 0;

                if (m_format.sampleSize() == 8
                    && m_format.sampleType() == QAudioFormat::UnSignedInt) {
                    value = *reinterpret_cast<const quint8*>(ptr);
                } else if (m_format.sampleSize() == 8
                           && m_format.sampleType() == QAudioFormat::SignedInt) {
                    value = qAbs(*reinterpret_cast<const qint8*>(ptr));
                } else if (m_format.sampleSize() == 16
                           && m_format.sampleType() == QAudioFormat::UnSignedInt) {
                    if (m_format.byteOrder() == QAudioFormat::LittleEndian)
                        value = qFromLittleEndian<quint16>(ptr);
                    else
                        value = qFromBigEndian<quint16>(ptr);
                } else if (m_format.sampleSize() == 16
                           && m_format.sampleType() == QAudioFormat::SignedInt) {
                    if (m_format.byteOrder() == QAudioFormat::LittleEndian)
                        value = qAbs(qFromLittleEndian<qint16>(ptr));
                    else
                        value = qAbs(qFromBigEndian<qint16>(ptr));
                } else if (m_format.sampleSize() == 32
                           && m_format.sampleType() == QAudioFormat::UnSignedInt) {
                    if (m_format.byteOrder() == QAudioFormat::LittleEndian)
                        value = qFromLittleEndian<quint32>(ptr);
                    else
                        value = qFromBigEndian<quint32>(ptr);
                } else if (m_format.sampleSize() == 32
                           && m_format.sampleType() == QAudioFormat::SignedInt) {
                    if (m_format.byteOrder() == QAudioFormat::LittleEndian)
                        value = qAbs(qFromLittleEndian<qint32>(ptr));
                    else
                        value = qAbs(qFromBigEndian<qint32>(ptr));
                } else if (m_format.sampleSize() == 32
                           && m_format.sampleType() == QAudioFormat::Float) {
                    value = qAbs(*reinterpret_cast<const float*>(ptr) * 0x7fffffff); // assumes 0-1.0
                }

                maxValue = qMax(value, maxValue);
                ptr += channelBytes;
            }
        }

        maxValue = qMin(maxValue, m_maxAmplitude);
        m_level = qreal(maxValue) / m_maxAmplitude;
    }

    emit update();
    return len;
}

QtAudioCapturer::QtAudioCapturer(QObject* parent)
    : QObject(parent)
    , m_deviceType(Microphone)
    , m_isRunning(false)
    , m_audioDevice(nullptr)
{
    // 初始化缓冲区
    m_buffer.open(QIODevice::ReadWrite);

    // 设置处理定时器
    connect(&m_processTimer, &QTimer::timeout, this, &QtAudioCapturer::processAudioData);
}

QtAudioCapturer::~QtAudioCapturer()
{
    Stop();
}

bool QtAudioCapturer::Init(Type deviceType)
{
    m_deviceType = deviceType;
    setupAudioFormat();
    if (m_deviceType == Microphone) {
        return initMicrophone();
    } else {
        return initSpeaker();
    }
}

bool QtAudioCapturer::Start()
{
    if (m_isRunning) {
        qDebug() << "QtAudioCapturer::Start - 已经在运行中";
        return true;
    }

    if (!m_audioInput && m_deviceType == Microphone) {
        qWarning() << "QtAudioCapturer::Start - 音频输入未初始化";
        return false;
    }

    qDebug() << "QtAudioCapturer::Start - 开始捕获音频，设备类型:"
             << (m_deviceType == Microphone ? "麦克风" : "扬声器");

    if (m_deviceType == Microphone && m_audioInput) {
        m_audioDevice = m_audioInput->start();
        if (!m_audioDevice) {
            qWarning() << "QtAudioCapturer::Start - 启动音频输入失败";
            return false;
        }
        qDebug() << "QtAudioCapturer::Start - QAudioInput state:" << m_audioInput->state();
        qDebug() << "QtAudioCapturer::Start - QAudioInput error:" << m_audioInput->error();
        connect(m_audioDevice, &QIODevice::readyRead, this, &QtAudioCapturer::handleReadyRead);
    }
    m_isRunning = true;
    return true;
}

void QtAudioCapturer::Stop()
{
    if (!m_isRunning) {
        return;
    }

    m_processTimer.stop();

    if (m_audioDevice) {
        disconnect(m_audioDevice, &QIODevice::readyRead, this, &QtAudioCapturer::handleReadyRead);
        m_audioDevice = nullptr;
    }

    if (m_audioInput) {
        m_audioInput->stop();
    }

    m_buffer.buffer().clear();
    m_buffer.seek(0);
    m_isRunning = false;
}

const AudioFormat& QtAudioCapturer::GetFormat() const
{
    return m_audioFormat;
}

void QtAudioCapturer::handleReadyRead()
{
    //qDebug() << "handleReadyRead called, m_audioDevice:" << m_audioDevice;
    if (m_audioDevice) {
        QByteArray data = m_audioDevice->readAll();
        //qDebug() << "handleReadyRead: read data size:" << data.size();
        if (data.size() > 0) {
            QMutexLocker locker(&m_mutex);
            m_dataBuffer.append(data);
            //qDebug() << "Audio data appended, buffer size now:" << m_dataBuffer.size();
        }
    }
}

int QtAudioCapturer::readAudioData(char* buf, int maxLen)
{
    QMutexLocker locker(&m_mutex);
    int toRead = qMin(maxLen, m_dataBuffer.size());
    // qDebug() << "readAudioData called, buffer size:" << m_dataBuffer.size() << ", toRead:" << toRead;
    if (toRead > 0) {
        memcpy(buf, m_dataBuffer.constData(), toRead);
        m_dataBuffer.remove(0, toRead);
    }
    return toRead;
}

void QtAudioCapturer::processAudioData()
{
    QByteArray data = m_buffer.buffer();
    if (data.size() > 0) {
        // qDebug() << "QtAudioCapturer::processAudioData - 处理缓冲区数据，大小:" << data.size()
        //          << "字节";

        // 清空缓冲区
        m_buffer.buffer().clear();
        m_buffer.seek(0);
    }
}

void QtAudioCapturer::onAudioNotify()
{
    if (m_audioDevice && m_audioDevice->bytesAvailable() > 0) {
        handleReadyRead();
    } else {
        // 尝试直接从 QAudioInput 读取数据
        QByteArray buffer(m_audioInput->bufferSize() / 4, 0);
        qint64 len = buffer.size();

        if (len > 0) {
            qDebug() << "QtAudioCapturer::onAudioNotify - 直接读取到数据，大小:" << len << "字节";
            buffer.resize(len);

            // 调用回调函数
            // if (m_callback) {
            //     m_callback(buffer.data(), buffer.size(), m_userInfo);
            // }
        }
    }
}

bool QtAudioCapturer::initMicrophone()
{
    // 获取默认输入设备
    QAudioDeviceInfo deviceInfo = QAudioDeviceInfo::defaultInputDevice();

    qDebug() << "QtAudioCapturer::initMicrophone - 默认输入设备:" << deviceInfo.deviceName();

    // 检查设备是否有效
    if (deviceInfo.isNull()) {
        qWarning() << "QtAudioCapturer::initMicrophone - 没有可用的音频输入设备";
        return false;
    }

    // 列出所有可用的音频输入设备
    qDebug() << "QtAudioCapturer::initMicrophone - 可用的音频输入设备:";
    for (const auto& device : QAudioDeviceInfo::availableDevices(QAudio::AudioInput)) {
        qDebug() << "  " << device.deviceName();
    }

    // 检查格式是否支持
    if (!deviceInfo.isFormatSupported(m_qtAudioFormat)) {
        qWarning() << "QtAudioCapturer::initMicrophone - 默认格式不支持，尝试使用最接近的格式";

        // 尝试几种常用格式
        QList<int> sampleRates = {44100, 48000, 16000, 8000};
        QList<int> channelCounts = {2, 1};
        QList<int> sampleSizes = {16, 8};

        bool formatFound = false;

        for (int rate : sampleRates) {
            for (int channels : channelCounts) {
                for (int size : sampleSizes) {
                    QAudioFormat format;
                    format.setSampleRate(rate);
                    format.setChannelCount(channels);
                    format.setSampleSize(size);
                    format.setCodec("audio/pcm");
                    format.setByteOrder(QAudioFormat::LittleEndian);
                    format.setSampleType(QAudioFormat::SignedInt);

                    if (deviceInfo.isFormatSupported(format)) {
                        m_qtAudioFormat = format;
                        formatFound = true;
                        qDebug() << "QtAudioCapturer::initMicrophone - 找到支持的格式:"
                                 << "采样率=" << rate << "通道数=" << channels
                                 << "采样大小=" << size;
                        break;
                    }
                }
                if (formatFound)
                    break;
            }
            if (formatFound)
                break;
        }

        if (!formatFound) {
            // 使用设备支持的最接近的格式
            m_qtAudioFormat = deviceInfo.nearestFormat(m_qtAudioFormat);
            qDebug() << "QtAudioCapturer::initMicrophone - 使用最接近的格式:"
                     << "采样率=" << m_qtAudioFormat.sampleRate()
                     << "通道数=" << m_qtAudioFormat.channelCount()
                     << "采样大小=" << m_qtAudioFormat.sampleSize();
        }

        // 更新 AudioFormat
        m_audioFormat = AudioFormat(m_qtAudioFormat.sampleRate(),
                                    m_qtAudioFormat.channelCount(),
                                    m_qtAudioFormat.sampleSize());
    }

    // 创建音频输入
    m_audioInput = std::make_unique<QAudioInput>(deviceInfo, m_qtAudioFormat, this);

    m_audioInfo.reset(new AudioInfo(m_qtAudioFormat));
    m_audioInfo->start();
    qDebug() << "QtAudioCapturer::initMicrophone - 音频输入创建成功，格式:"
             << "采样率=" << m_qtAudioFormat.sampleRate()
             << "通道数=" << m_qtAudioFormat.channelCount()
             << "采样大小=" << m_qtAudioFormat.sampleSize();

    return true;
}

bool QtAudioCapturer::initSpeaker()
{
    return false;
}

void QtAudioCapturer::setupAudioFormat()
{
    // 设置Qt音频格式
    m_qtAudioFormat.setSampleRate(44100); // 与原代码中的 DEFAULT_SAMPLE_RATE 保持一致
    m_qtAudioFormat.setChannelCount(1);   // 改为单声道，提高兼容性
    m_qtAudioFormat.setSampleSize(16);    // 与原代码中的 DEFAULT_BITS_PER_SAMPLE 保持一致
    m_qtAudioFormat.setCodec("audio/pcm");
    m_qtAudioFormat.setByteOrder(QAudioFormat::LittleEndian);
    m_qtAudioFormat.setSampleType(QAudioFormat::SignedInt);

    // 设置IAudioCapturer使用的格式
    m_audioFormat = AudioFormat(m_qtAudioFormat.sampleRate(),
                                m_qtAudioFormat.channelCount(),
                                m_qtAudioFormat.sampleSize());
}

float QtAudioCapturer::calculateVolume(const QByteArray& data)
{
    // 计算音频数据的音量级别
    if (data.isEmpty()) {
        return 0.0f;
    }

    // 假设数据是16位有符号整数
    const int16_t* samples = reinterpret_cast<const int16_t*>(data.constData());
    int sampleCount = data.size() / sizeof(int16_t);

    // 计算平均绝对值
    float sum = 0.0f;
    for (int i = 0; i < sampleCount; ++i) {
        sum += std::abs(samples[i]);
    }

    // 归一化到0-1范围
    float avgAmp = sum / (sampleCount * 32768.0f);
    return avgAmp;
}