StemCollection.h

   1 // Mini-lexicon’s collection of stems
   2 // Copyright © 2009 The University of Chicago
   3 #ifndef STEMCOLLECTION_H
   4 #define STEMCOLLECTION_H
   5
   6 // See the CMiniLexicon class in MiniLexicon.h for an overview of
   7 // suffix/signature-based discovery of morphology.
   8
   9 class CStemCollection;
  10
  11 #include "CollectionTemplate.h"
  12 #include "AffixLocation.h"
  13 #include "generaldefinitions.h"
  14 template<class K, class V> class QMap;
  15
  16 /// List of stems for a stem/signature model of morphology
  17 ///
  18 /// Maintains a list of stems used in analyzing a particular collection
  19 /// of words (that is, all stems associated to a particular mini-lexicon).
  20 /// Semantically, a stem collection is a set (like std::set) of pointers
  21 /// to CStem objects which it owns.  The contained stems are generally
  22 /// deleted upon removal.
  23 ///
  24 /// When a stem is inserted, it is added to the associated lexicon, and
  25 /// when a stem is removed, it is removed from the associated lexicon.
  26 ///
  27 /// This collection type also maintains a total use count for the stems it
  28 /// manages, for use in description length calculations.
  29 ///
  30 /// To support the successor-frequency algorithm, a stem collection
  31 /// should support incremental lookup of stems (find all stems starting
  32 /// with a given phoneme sequence, add a phoneme to that sequence to
  33 /// narrow the search, etc).
  34 class CStemCollection : public TCollection<class CStem> {
  35 protected:
  36         enum eAffixLocation m_AffixLocation;
  37         double m_TotalUseCount;
  38 public:
  39         // construction/destruction.
  40
  41         CStemCollection();
  42         CStemCollection(class CMiniLexicon* mini);
  43         ~CStemCollection();
  44         // copy construction, copy-assignment implicitly defined.
  45
  46         // Qt3-style collection view.
  47
  48         void ListDisplay(class Q3ListView* parent,
  49                 QMap<class QString, class QString>* filter = 0);
  50
  51         // input/output to file.
  52
  53         void OutputStems(class QString filename,
  54                 QMap<class QString, class QString>* filter);
  55         void ReadStemFile(class QString filename,
  56                         enum eAffixLocation affix_loc);
  57
  58         // insert.
  59
  60         CStem* operator<<(const CStem* stem);
  61         CStem* operator<<(class CStringSurrogate stem_text);
  62         CStem* operator<<(const class CParse* stem_text);
  63         CStem* operator<<(class QString stem_text);
  64         void AddPointer(CStem* pointee);
  65         CStem* AddToCollection(const class CParse& stem_text);
  66         CStem* AddToCollection(const class CStringSurrogate& stem_text);
  67
  68         // clear.
  69
  70         void Empty();
  71         void RemoveAll();
  72
  73         // remove.
  74
  75         bool Remove(CStem* stem); // doesn't delete CStem*
  76         bool RemoveMember(CStem* stem); // deletes CStem*
  77         bool RemoveMember(const class CStringSurrogate& stem_text); // deletes CStem*
  78         bool RemoveMember(const class CStringSurrogate& stem_text,
  79                                 bool delete_stem); // FSA
  80         void DeleteMarkedMembers();
  81
  82         // accessors.
  83
  84         double GetTotalUseCount();
  85
  86         // description length.
  87
  88         double CalculateSumOfPointersToMyStems(enum eMDL_STYLE style);
  89         double CalculateTotalPhonologicalInformationContent(
  90                 class CLexicon* MotherLexicon);
  91 };
  92
  93 #endif // STEMCOLLECTION_H