File indexing completed on 2024-05-12 15:55:08

0001 /************************************************************************
0002  *                                  *
0003  *  This file is part of Kooka, a scanning/OCR application using    *
0004  *  Qt <http://www.qt.io> and KDE Frameworks <http://www.kde.org>.  *
0005  *                                  *
0006  *  Copyright (C) 2020      Jonathan Marten <jjm@keelhaul.me.uk>    *
0007  *                                  *
0008  *  Kooka is free software; you can redistribute it and/or modify it    *
0009  *  under the terms of the GNU Library General Public License as    *
0010  *  published by the Free Software Foundation and appearing in the  *
0011  *  file COPYING included in the packaging of this file;  either    *
0012  *  version 2 of the License, or (at your option) any later version.    *
0013  *                                  *
0014  *  As a special exception, permission is given to link this program    *
0015  *  with any version of the KADMOS OCR/ICR engine (a product of     *
0016  *  reRecognition GmbH, Kreuzlingen), and distribute the resulting  *
0017  *  executable without including the source code for KADMOS in the  *
0018  *  source distribution.                        *
0019  *                                  *
0020  *  This program is distributed in the hope that it will be useful, *
0021  *  but WITHOUT ANY WARRANTY; without even the implied warranty of  *
0022  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the   *
0023  *  GNU General Public License for more details.            *
0024  *                                  *
0025  *  You should have received a copy of the GNU General Public       *
0026  *  License along with this program;  see the file COPYING.  If     *
0027  *  not, see <http://www.gnu.org/licenses/>.                *
0028  *                                  *
0029  ************************************************************************/
0030 
0031 #ifndef OCRTESSERACTENGINE_H
0032 #define OCRTESSERACTENGINE_H
0033 
0034 #include "abstractocrengine.h"
0035 
0036 
0037 class OcrTesseractEngine : public AbstractOcrEngine
0038 {
0039     Q_OBJECT
0040 
0041 public:
0042     explicit OcrTesseractEngine(QObject *pnt, const QVariantList &args);
0043     ~OcrTesseractEngine() override = default;
0044 
0045     AbstractOcrDialogue *createOcrDialogue(AbstractOcrEngine *plugin, QWidget *pnt) override;
0046 
0047     bool hasAdvancedSettings() const override       { return (true); }
0048     void openAdvancedSettings() override;
0049 
0050 protected:
0051     bool createOcrProcess(AbstractOcrDialogue *dia, ScanImage::Ptr img) override;
0052     QStringList tempFiles(bool retain) override;
0053     bool finishedOcrProcess(QProcess *proc) override;
0054 
0055 private:
0056     QString readHOCR(const QString &fileName);
0057 
0058 private:
0059     QString m_ocrImageIn;
0060     QString m_tempHOCROut;
0061     QString m_tempStdoutLog;
0062 
0063     int m_tesseractVersion;
0064 };
0065 
0066 #endif                          // OCRTESSERACTENGINE_H