Mercurial > hg > sonic-annotator
view runner/LabFeatureWriter.cpp @ 399:a3912193ce69 tip
Default branch is now named default on git as well as hg, in case we ever want to switch to mirroring in the other direction
author | Chris Cannam |
---|---|
date | Thu, 27 Aug 2020 15:57:37 +0100 |
parents | e39307a8d22d |
children |
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. Sonic Annotator A utility for batch feature extraction from audio files. Mark Levy, Chris Sutton and Chris Cannam, Queen Mary, University of London. Copyright 2007-2008 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 "LabFeatureWriter.h" #include <iostream> #include <QRegExp> #include <QTextStream> #include <QTextCodec> using namespace std; using namespace Vamp; LabFeatureWriter::LabFeatureWriter() : FileFeatureWriter(SupportOneFilePerTrackTransform | SupportStdOut, "lab"), m_forceEnd(false), m_digits(6) { } LabFeatureWriter::~LabFeatureWriter() { } string LabFeatureWriter::getDescription() const { return "Write features in .lab, a tab-separated columnar format. The first column is always the feature start time in seconds. If the features have duration, the second column will be the feature end time in seconds. Remaining columns are the feature values (if any) and finally the feature label (if any). There is no identification of the audio file or the transform, so confusion will result if features from different audio or transforms are mixed. (For more control over the output, consider using the more general CSV writer.)"; } LabFeatureWriter::ParameterList LabFeatureWriter::getSupportedParameters() const { ParameterList pl = FileFeatureWriter::getSupportedParameters(); Parameter p; p.name = "fill-ends"; p.description = "Include end times even for features without duration, by using the gap to the next feature instead."; p.hasArg = false; pl.push_back(p); p.name = "digits"; p.description = "Specify the number of significant digits to use when printing transform outputs. Outputs are represented internally using single-precision floating-point, so digits beyond the 8th or 9th place are usually meaningless. The default is 6."; p.hasArg = true; pl.push_back(p); return pl; } void LabFeatureWriter::setParameters(map<string, string> ¶ms) { FileFeatureWriter::setParameters(params); for (map<string, string>::iterator i = params.begin(); i != params.end(); ++i) { if (i->first == "fill-ends") { m_forceEnd = true; } else if (i->first == "digits") { int digits = atoi(i->second.c_str()); if (digits <= 0 || digits > 100) { SVCERR << "LabFeatureWriter: ERROR: Invalid or out-of-range value for number of significant digits: " << i->second << endl; SVCERR << "LabFeatureWriter: NOTE: Continuing with default settings" << endl; } else { m_digits = digits; } } } } void LabFeatureWriter::write(QString trackId, const Transform &transform, const Plugin::OutputDescriptor& , const Plugin::FeatureList& features, std::string) { // Select appropriate output file for our track/transform // combination TransformId transformId = transform.getIdentifier(); QTextStream *sptr = getOutputStream (trackId, transformId, QTextCodec::codecForName("UTF-8")); if (!sptr) { throw FailedToOpenOutputStream(trackId, transformId); } QTextStream &stream = *sptr; int n = int(features.size()); if (n == 0) return; DataId tt(trackId, transform); if (m_pending.find(tt) != m_pending.end()) { writeFeature(stream, m_pending[tt], &features[0]); m_pending.erase(tt); } if (m_forceEnd) { // can't write final feature until we know its end time --n; m_pending[tt] = features[n]; } for (int i = 0; i < n; ++i) { writeFeature(stream, features[i], m_forceEnd ? &features[i+1] : 0); } } void LabFeatureWriter::finish() { for (PendingFeatures::const_iterator i = m_pending.begin(); i != m_pending.end(); ++i) { DataId tt = i->first; Plugin::Feature f = i->second; QTextStream *sptr = getOutputStream (tt.first, tt.second.getIdentifier(), QTextCodec::codecForName("UTF-8")); if (!sptr) { throw FailedToOpenOutputStream(tt.first, tt.second.getIdentifier()); } QTextStream &stream = *sptr; // final feature has its own time as end time (we can't // reliably determine the end of audio file, and because of // the nature of block processing, the feature could even // start beyond that anyway) writeFeature(stream, f, &f); } m_pending.clear(); } void LabFeatureWriter::writeFeature(QTextStream &stream, const Plugin::Feature &f, const Plugin::Feature *optionalNextFeature) { QString sep = "\t"; QString timestamp = f.timestamp.toString().c_str(); timestamp.replace(QRegExp("^ +"), ""); stream << timestamp; Vamp::RealTime endTime; bool haveEndTime = true; if (f.hasDuration) { endTime = f.timestamp + f.duration; } else if (optionalNextFeature) { endTime = optionalNextFeature->timestamp; } else { haveEndTime = false; } if (haveEndTime) { QString e = endTime.toString().c_str(); e.replace(QRegExp("^ +"), ""); stream << sep << e; } for (unsigned int j = 0; j < f.values.size(); ++j) { stream << sep << QString("%1").arg(f.values[j], 0, 'g', m_digits); } if (f.label != "") { stream << sep << "\"" << f.label.c_str() << "\""; } stream << "\n"; }