linguisticamainwindow.h

   1 // User interface and driver
   2 // Copyright © 2009 The University of Chicago
   3 #ifndef LINGUISTICAMAINWINDOW_H
   4 #define LINGUISTICAMAINWINDOW_H
   5
   6 // Doxygen comments defining the index page
   7 /**     \mainpage Linguistica
   8
   9         \section intro_sec What is Linguistica?
  10
  11         Linguistica is a program designed to explore the unsupervised learning of natural language, with primary focus on morphology (word-structure). It runs under Windows, Mac OS X and Linux, and is written in C++ within the Qt development framework. Its demands on memory depend on the size of the corpus analyzed.
  12
  13         Unsupervised learning refers to the computational task of making inferences (and therefore acquiring knowledge) about the structure that lies behind some set of data, without any direct access to that structure. In the case of unsupervised learning of morphology, Linguistica explores the possibilities of morpheme-combinations for a set of words, based on no internal knowledge of the language from which the words are drawn.
  14
  15         Segmentation is the first task of this process; the program figures out where the morpheme boundaries are in the words, and then decides what the stems are, what the suffixes are, and so forth. Most of Linguistica's functionality, at this point, goes into making these decisions.
  16
  17         \section creator Creator
  18         - John Goldsmith
  19
  20         \section engineers Engineers
  21
  22         - Jon Cihlar
  23         - Yu Hu
  24         - Jeremy O'Brien
  25         - Colin Sprague
  26
  27         \section contributors Contributors
  28
  29         - Carl Dolnick
  30         - James Kirby
  31         - Irina Matveeva
  32         - Aris Xanthos
  33
  34  */
  35
  36 #include <QMainWindow>
  37 #include "ui_linguisticamainwindowbase.h"
  38
  39 #include <memory>
  40 #include <Q3ProgressDialog>
  41 #include <Q3FileDialog>
  42 #include <QInputDialog>
  43 #include <QFontDialog>
  44 #include <Q3PopupMenu>
  45 #include <QStatusBar>
  46 #include <Q3TextEdit>
  47 #include <QSplitter>
  48 #include <Q3ToolBar>
  49 #include <QMenuBar>
  50 #include <Q3Canvas>
  51 #include <Q3VBox>
  52 #include <QResizeEvent>
  53 #include <QCloseEvent>
  54 #include <QMoveEvent>
  55 #include <QApplication>
  56 #include <QSettings>
  57 #include <Q3Process>
  58 #include <QLayout>
  59 #include <QAction>
  60 #include <QLabel>
  61 #include <Q3TextStream>
  62 #include <Q3PtrList>
  63 #include <QString>
  64 #include <Qt>
  65 #include "LPreferences.h"
  66 #include "generaldefinitions.h"
  67 #include "AffixLocation.h"
  68
  69 #include <QLabel>
  70
  71 namespace linguistica { namespace ui {
  72         struct status_user_agent;
  73 } }
  74 class CStem;
  75 class CLexicon;
  76 class CParse;
  77 class CLPreferences;
  78 class CTreeViewItem;
  79 class CState;
  80 class GraphicView;
  81 class CTemplate;
  82 class CTemplateCollection;
  83 class CTemplateListViewItem;
  84 class CSequencer;
  85 class CStats;
  86 class cMT;
  87 class CCorpusWord;
  88 class CSignature;
  89 class GUIWordCollection;
  90
  91 enum eDocumentType;
  92
  93 //typedef QMap<QString,QString>                 StringToString;
  94 typedef QMap<QString,CTemplate*>                QStringToTemplate;
  95 typedef Q3PtrList<CStem>                                CStemList;
  96 typedef QMap<QString,CStemList>                 StringToCStemList;
  97 typedef QMap<QString,CStem*>                    StringToPtrCStem;
  98 typedef QMap<QString,CStem>                     StringToCStem;
  99 typedef QMap<QString,CParse*>                   StringToParse;
 100 typedef QMap<QString,CCorpusWord*>              StringToPtrCCorpusWord;
 101 typedef QMap<int,QString>                       IntToString;
 102
 103 class LinguisticaMainWindow : public QMainWindow,
 104                                         public Ui::LinguisticaMainWindowBase
 105 {
 106         Q_OBJECT
 107 protected:
 108         // Application
 109         QApplication*          m_app;   ///< memory belongs to caller
 110         /// true if running in batch mode
 111         bool                   m_commandLineFlag;
 112
 113         // Preferences
 114         QSettings              m_Settings;
 115         CLPreferences          m_preferences;
 116
 117         // Lexicon
 118         CLexicon*              m_lexicon;
 119
 120         // GUI
 121         eDocumentType          m_docType;
 122         CParse*                m_commandParse;
 123         int                    m_commandMini;
 124         CSignature*            m_pSig1;
 125         CSignature*            m_pSig2;
 126         GUIWordCollection*     m_GUIWords;
 127
 128         // Token count
 129         int                    m_numberOfTokens;
 130
 131         // Logging variables
 132         bool                   m_logging;
 133         QString                m_logFileName;
 134         QFile*                 m_logFile;
 135         QTextStream*           m_logFileStream;
 136
 137         // Status/Progress Bar
 138         std::auto_ptr<linguistica::ui::status_user_agent> m_status;
 139         Q3ProgressDialog*     m_progressDialog;
 140         QString               m_progressText1,
 141                               m_progressText2;
 142
 143         // Counter
 144         bool                  m_counterDisplayFlag;
 145         int                   m_counter;
 146
 147         // Project information
 148         int                  m_projectIndex;
 149         QString              m_projectDirectory;
 150         QString              m_projectName;
 151         bool                 m_projectDirty;
 152
 153         // File names
 154         QString              m_corpusFileName;
 155         QString              m_goldStdFileName;
 156         QString              m_goldStdOutputFile;
 157
 158         // Encoding
 159         QString m_EncodingType;
 160
 161         // Fonts
 162         // Collection view and command line can have different
 163         // font, they are on the east side of the screen
 164         QFont                m_eastFont;  // Collection view and command line font, east side of screen
 165
 166
 167         //========== FSA
 168         CState*              m_CommonState;
 169
 170         //========= SED
 171         CTemplate*           m_CommonTemplate;
 172         CTemplateCollection* m_Words_InitialTemplates;
 173         CTemplateCollection* m_Words_Templates;
 174
 175         //========= Sequencer
 176         CSequencer*          m_Sequencer;
 177
 178         //========= Statistics
 179         CStats*              m_Stats;
 180
 181         //========= MT
 182         cMT*                 m_MT;
 183
 184 public:
 185   LinguisticaMainWindow( const int argc, char** argv, QApplication* app = 0,
 186                   QWidget* parent = 0, Qt::WindowFlags f = Qt::WType_TopLevel );
 187   ~LinguisticaMainWindow();
 188
 189         // disable copy
 190 private:
 191         LinguisticaMainWindow(const LinguisticaMainWindow& x);
 192         LinguisticaMainWindow& operator=(const LinguisticaMainWindow& x);
 193 public:
 194
 195   void                       SetNumberOfTokens( int i ) { m_numberOfTokens = i; }
 196   int                        GetNumberOfTokens() { return m_numberOfTokens; }
 197
 198   void                       SetUpNewLexicon    ();
 199
 200   void                       setProjectDirectory( QString dir ) { m_projectDirectory = dir; }
 201   bool                                  getProjectDirty () { return m_projectDirty; }
 202
 203   void                                  setStatusBar1   (QString);
 204   void                                  setStatusBar2   (QString);
 205
 206   void                                  BeginCountDown  ();
 207   void                                  EndCountDown    ();
 208   void                                  CountDownOnStatusBar(int i, int total, int step = 100);
 209
 210         linguistica::ui::status_user_agent& status_display()
 211                 { return *m_status; }
 212
 213   QString                               GetLogFileName  () { return m_logFileName; }
 214   void                                  AskUserForLogFileName();
 215
 216   void                                  SetEastFont             ( QFont font ) { m_eastFont = font; }
 217   CLPreferences*                        GetPreferences  () { return &m_preferences; }
 218   QSettings*                            GetSettings             () { return &m_Settings; }
 219
 220   void                                  SetDocType              ( eDocumentType dt ) { m_docType = dt; }
 221   eDocumentType                         GetDocType              () { return m_docType; }
 222
 223   bool                                  autoLayering    () { return m_autoLayeringAction->isOn(); }
 224
 225   bool                                  isCorrectAffix  ( eAffixLocation );
 226
 227         void executeCommandLineArgs(int words_count, bool auto_layer,
 228                 bool prefix, bool suffix);
 229         /// true if running in batch mode
 230         bool commandLineMode() { return m_commandLineFlag; }
 231
 232   QFont                                 GetEastFont             () { return m_eastFont; }
 233
 234         // Logging
 235   void                                  CloseLogFile    ();
 236   void                                  OpenLogFile             ();
 237   bool                                  LogFileOn               () { return m_logging; }
 238   QTextStream*                          GetLogFileStream() { return m_logFileStream; }
 239
 240
 241         Q3Canvas*       m_SmallCanvas;
 242         GraphicView*    m_SmallGraphicDisplay;  ///< deleted by m_SmallGraphicDisplayTab
 243         Q3VBox*         m_SmallGraphicDisplayVBox;
 244         QLabel * pImgLabel;
 245
 246         CLexicon*       GetLexicon(void) { return m_lexicon; }
 247 private:
 248
 249
 250   Q3ListViewItem*        GetMiniLexiconSubTree( Q3ListViewItem*, int );
 251   void                                  GetMorphPrecisionRecallByWord( StringToCStemList&,
 252                                                                                                            StringToParse&,
 253                                                                                                            double&,
 254                                                                                                            double&,
 255                                                                                                            double&,
 256                                                                                                            double& );
 257
 258   void                                  GetCutPrecisionRecall( StringToCStemList&,
 259                                                                                            StringToParse&,
 260                                                                                            double&,
 261                                                                                            double&,
 262                                                                                            double&,
 263                                                                                            double& );
 264
 265   void                                  GetMorphPrecisionRecall( StringToCStemList&,
 266                                                                                                  StringToParse&,
 267                                                                                                  double&,
 268                                                                                                  double&,
 269                                                                                                  double&,
 270                                                                                                  double& );
 271 private:
 272         QFont GetFontPreference(QString key) const
 273         { return m_preferences.GetFontPreference(key); }
 274         void SetFontPreference(QString key, QFont font)
 275         { m_preferences.SetFontPreference(key, font); }
 276         QString GetPreference(QString key) const
 277         { return m_preferences.GetPreference(key); }
 278         void GetStringListPreference(QString key, QStringList* out) const
 279         { m_preferences.GetStringListPreference(key, out); }
 280
 281 signals:
 282   void corpusOKSignal(bool);
 283   void updateAllViewsSignal();
 284   void updateTreeViewSignal();
 285   void updateCommandLineSignal();
 286   void tokensRequestedSignal();  // for changing the number of words requested
 287   void commandLineClose();
 288
 289 public slots:
 290
 291   // Slots derived from LinguisticaMainWindowBase
 292   void openFileActionSlot();
 293   void runAllSuffixActionSlot();
 294   void enableLoggingActionSlot( bool );
 295   void clearLogFileActionSlot();
 296   void newLogFileActionSlot();
 297   /// Refresh collection view, for example for a switch in m_docType.
 298   void updateCollectionViewSlot();
 299   void updateCollectionViewSlot( Q3ListViewItem* );
 300   void updateCommandLineSlot( Q3ListViewItem* );
 301   void changeFontSlot();
 302   void preferencesEditSlot();
 303   void saveFileActionSlot();
 304   void newFileActionSlot();
 305   void saveAsFileActionSlot();
 306   void findAllomorphySlot();
 307   void showDeleteesSlot( bool );
 308   void rereadCorpusSlot();
 309   void clearLexiconSlot();
 310   void predecessorFreqSlot();
 311   void successorFreqSlot();
 312   void checkPfxSignaturesSlot();
 313   void checkSignaturesSlot();
 314   void knownStemsPrefixesSlot();
 315   void knownStemsSuffixesSlot();
 316   void fromPfxSigsFindStemsSlot();
 317   void fromSigsFindStemsSlot();
 318   void findPfxSingletonSigsSlot();
 319   void findSingletonSigsSlot();
 320   void pfxLooseFitSlot();
 321   void looseFitSlot();
 322   void showToolbarsSlot( bool );
 323   void showStatusBarSlot( bool );
 324   void saveBrokenCorpusSlot();
 325   void helpAboutSlot();
 326   void newMiniLexiconActionSlot();
 327   void deleteMiniLexiconActionSlot();
 328   void autoLayeringActionSlot();
 329   void populateWordsActionSlot();
 330   void addToBorrowedSigsSlot();
 331   void compareGoldStdSlot();
 332   void changeGoldStdFileSlot();
 333   void runAllPrefixActionSlot();
 334   void commandLineCloseSlot();
 335   void viewCorpusSlot();
 336   void exportProjectSlot();
 337   void addContentToCommandLine(QString);
 338   void setCommandLineTabStopWidth(int length){m_commandLine->setTabStopWidth(length);};
 339   int  getCommandLinePointSize(){ return m_commandLine ->pointSize();};
 340   void setCommandLineStyle( QString );
 341   void xfstExportActionSlot();
 342
 343   // Compounds
 344   void fromStemsFindCompoundsSlot();
 345   void fromStemsFindFlatCompoundsSlot();
 346   void calculateAffixnessSlot();
 347   void fromAffixnessUpdateSigsAndCompoundsSlot();
 348   void compareCompoundsSlot();
 349
 350   // Phonology
 351   virtual void GenericPhonologySlot();
 352   virtual void CreateSkeletonSlot();
 353   virtual void fromStemsFindAffixesSlot();
 354   virtual void DropTierLeaveSlot();
 355   virtual void DropTierLeaveCopy();
 356   virtual void PhonologyCalculateInformation();
 357   virtual void PhonologyComputeGibbsFunction();
 358
 359   void updateSmallGraphicDisplaySlot(); //needed for compound displays
 360 // --- new fsa code
 361   void fsaTestFuncAction();
 362
 363   // Phonogy graphic display
 364   void updateSmallGraphicDisplaySlotForPhonogy(CStem*);
 365
 366
 367   // Sequencer
 368   void sequencerReadCorpusSlot();
 369   void sequenceASentenceSlot();
 370   void sequencerTestAFileSlot();
 371
 372   //========  MT
 373   void mTModel1Slot();
 374
 375    // ========= HMM
 376   void testSEMHMMSlot();
 377   void DisplayVideoHMM();
 378
 379   // ========= Graphic display of Multidimension data
 380   void testMultiDimensionDisplaySlot();
 381   void updateSmallGraphicDisplaySlotForMultiDimensionData(int, int, double**, IntToString);
 382
 383         // ======  Goldwater Method
 384   void goldwaterMethodSlot();
 385
 386   void testSignatureDisplaySlot(); //jg july 2006
 387
 388   // Statistics
 389   void statsReadCorpusSlot();
 390
 391   // DCN
 392   void DCNstressSlot();
 393   void DCNsylSlot();
 394
 395   void DCNstressComputeSlot();
 396   void DCNstressLearnSlot();
 397
 398   void DCNsylFindTrainSlot();
 399   void DCNsylFindTestSlot();
 400   void DCNsylTrainSlot();
 401
 402   // StringEditDistance
 403   void RunSedSlot();
 404   void FindSlicesSlot();
 405   void FindAllEditDistances(CTemplateCollection*);
 406   void CheckForConflations(CTemplateCollection* , CTemplateCollection* );
 407   void FindAllWordNeighbors(CLexicon*);
 408
 409   // Goldstandard
 410 //  void GetPrecisionRecall(StringToPtrCStem&, StringToParse&, double& , double& , double& , double& );
 411
 412   // Machine Translation
 413   void editCorpusForMT();
 414
 415   // Other slots
 416   void jgWebLinksActionSlot();
 417   void lxaWebLinksActionSlot();
 418   void updateTreeViewSlot();
 419   void updateCommandLineSlot();
 420   void tokensRequestedDialogSlot();
 421   void setProjectDirtySlot( bool b ) { m_projectDirty = b; }
 422
 423   void suffixesCollapse_two_signaturesAction_activated();
 424   void updateSignatureGraphicDisplaySlot();
 425
 426
 427   //    User interface items
 428   void  toggleMultipleSelectionSlot();
 429
 430
 431 protected:
 432   virtual void closeEvent( QCloseEvent * );
 433   virtual void resizeEvent( QResizeEvent * );
 434   virtual void moveEvent( QMoveEvent * );
 435
 436 public:
 437         bool    getLogging(){ return m_logging; }
 438 };
 439
 440 #endif // LINGUISTICAMAINWINDOW_H