view data/midi/MIDIInput.cpp @ 1211:5a1198083d9a piper

Pull out model creation into the transformer thread run(), so that all communications with the plugin server happen on a single thread. Then make the model accessor wait for them to be created (which still happens right at the start of processing) before returning.
author Chris Cannam
date Mon, 17 Oct 2016 14:18:23 +0100
parents e802e550a1f2
children 6b847a59d908
line wrap: on
line source
/* -*- c-basic-offset: 4 indent-tabs-mode: nil -*-  vi:set ts=8 sts=4 sw=4: */

/*
    Sonic Visualiser
    An audio file viewer and annotation editor.
    Centre for Digital Music, Queen Mary, University of London.
    This file copyright 2006-2009 Chris Cannam and QMUL.
    
    This program is free software; you can redistribute it and/or
    modify it under the terms of the GNU General Public License as
    published by the Free Software Foundation; either version 2 of the
    License, or (at your option) any later version.  See the file
    COPYING included with this distribution for more information.
*/

#include "MIDIInput.h"

#include "rtmidi/RtMidi.h"

#include <unistd.h>

MIDIInput::MIDIInput(QString name, FrameTimer *timer) :
    m_rtmidi(),
    m_frameTimer(timer),
    m_buffer(1023)
{
    try {
        m_rtmidi = new RtMidiIn(name.toStdString());
        m_rtmidi->setCallback(staticCallback, this);
        m_rtmidi->openPort(0, tr("Input").toStdString());
    } catch (RtError e) {
        e.printMessage();
        delete m_rtmidi;
        m_rtmidi = 0;
    }
}

MIDIInput::~MIDIInput()
{
    delete m_rtmidi;
}

void
MIDIInput::staticCallback(double timestamp, std::vector<unsigned char> *message,
                          void *userData)
{
    ((MIDIInput *)userData)->callback(timestamp, message);
}

void
MIDIInput::callback(double timestamp, std::vector<unsigned char> *message)
{
    SVDEBUG << "MIDIInput::callback(" << timestamp << ")" << endl;
    // In my experience so far, the timings passed to this function
    // are not reliable enough to use.  We request instead an audio
    // frame time from whatever FrameTimer we have been given, and use
    // that as the event time.
    if (!message || message->empty()) return;
    unsigned long t = m_frameTimer->getFrame();
    MIDIByte code = (*message)[0];
    MIDIEvent ev(t,
                 code,
                 message->size() > 1 ? (*message)[1] : 0,
                 message->size() > 2 ? (*message)[2] : 0);
    postEvent(ev);
}

MIDIEvent
MIDIInput::readEvent()
{
    MIDIEvent *event = m_buffer.readOne();
    MIDIEvent revent = *event;
    delete event;
    return revent;
}

void
MIDIInput::postEvent(MIDIEvent e)
{
    int count = 0, max = 5;
    while (m_buffer.getWriteSpace() == 0) {
        if (count == max) {
            cerr << "ERROR: MIDIInput::postEvent: MIDI event queue is full and not clearing -- abandoning incoming event" << endl;
            return;
        }
        cerr << "WARNING: MIDIInput::postEvent: MIDI event queue (capacity " << m_buffer.getSize() << " is full!" << endl;
        SVDEBUG << "Waiting for something to be processed" << endl;
#ifdef _WIN32
        Sleep(1);
#else
        sleep(1);
#endif
        count++;
    }

    MIDIEvent *me = new MIDIEvent(e);
    m_buffer.write(&me, 1);
    emit eventsAvailable();
}