kiten/dict.h

   1 /**
   2  This file is part of Kiten, a KDE Japanese Reference Tool...
   3  Copyright (C) 2001  Jason Katz-Brown <jason@katzbrown.com>
   4
   5  This program is free software; you can redistribute it and/or modify
   6  it under the terms of the GNU General Public License as published by
   7  the Free Software Foundation; either version 2 of the License, or
   8  (at your option) any later version.
   9
  10  This program is distributed in the hope that it will be useful,
  11  but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13  GNU General Public License for more details.
  14
  15  You should have received a copy of the GNU General Public License
  16  along with this program; if not, write to the Free Software
  17  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307
  18  USA
  19 **/
  20
  21
  22 #ifndef DICT_H
  23 #define DICT_H
  24
  25 #include <qfile.h>
  26 #include <qmemarray.h>
  27 #include <qptrlist.h>
  28 #include <qregexp.h>
  29 #include <qstringlist.h>
  30
  31 #include <sys/types.h>
  32 #ifdef __osf__
  33 typedef unsigned int uint32_t;
  34 typedef int int32_t;
  35 #else
  36 #include <inttypes.h>
  37 #endif
  38
  39 namespace Dict
  40 {
  41
  42 enum TextType { Text_Kanji, Text_Kana, Text_Latin };
  43
  44 // returns the TextType of the first part of the text
  45 KDE_EXPORT TextType textType(const QString &text);
  46
  47 // File needs to be able to give out Arrays based on its mmap'd data.
  48 // But, we don't want the users of the arrays to have to remember to
  49 // resetRawData() after using them, since that's bound to fail sooner or later.
  50 //
  51 // This class handles it for us.
  52 template<class T> class Array : public QMemArray<T>
  53 {
  54 public:
  55         Array(T *, int);
  56         virtual ~Array();
  57
  58 private:
  59         T *data;
  60         int dataSize;
  61 };
  62
  63 template<class T> Array<T>::Array(T *d, int s)
  64         : QMemArray<T>()
  65         , data(d)
  66         , dataSize(s)
  67 {
  68         setRawData(data, dataSize / sizeof(T));
  69 }
  70
  71 template<class T> Array<T>::~Array()
  72 {
  73         resetRawData(data, dataSize / sizeof(T));
  74 }
  75
  76 // File manages all the files, pointers, and memory management associated
  77 // with a single dictionary.
  78 class File
  79 {
  80 public:
  81         File(QString path, QString name);
  82         ~File();
  83
  84         QString name(void);
  85
  86         Array<const unsigned char> dict(void);
  87         Array<const uint32_t> index(void);
  88
  89         int dictLength(void);
  90         int indexLength(void);
  91
  92         // replacement for exceptions thrown in the constructor
  93         bool isValid(void);
  94
  95         unsigned char lookup(unsigned i, int offset);
  96         QCString lookup(unsigned i);
  97 private:
  98         QString myName;
  99
 100         QFile dictFile;
 101         const unsigned char * dictPtr;
 102
 103         QFile indexFile;
 104         const uint32_t * indexPtr;
 105
 106         bool valid;
 107 };
 108
 109 class KDE_EXPORT Entry
 110 {
 111 public:
 112         // EDict ctor
 113         Entry(const QString &, const QString &, const QStringList &);
 114         // Kanjidict ctor
 115         Entry(const QString &, QStringList &, QStringList &, unsigned int grade, unsigned int freq, unsigned int strokes, unsigned int miscount);
 116         // default (for containers)
 117         Entry(const QString & = QString::null);
 118         // for a heading
 119         Entry(const QString &, bool header);
 120
 121         QString dictName();
 122         QString header();
 123         QStringList meanings();
 124         QStringList readings();
 125         QString firstReading();
 126
 127         bool kanaOnly();
 128         QString kanji();
 129
 130         bool extendedKanjiInfo();
 131         unsigned int grade();
 132         unsigned int strokes();
 133         unsigned int miscount();
 134         unsigned int freq();
 135
 136 protected:
 137         QString DictName;
 138         QString Header;
 139         QStringList Meanings;
 140
 141         QString Kanji;
 142         bool KanaOnly;
 143         QStringList Readings;
 144
 145         bool ExtendedKanjiInfo;
 146         unsigned int Grade;
 147         unsigned int Strokes;
 148         unsigned int Miscount;
 149         unsigned int Freq;
 150 };
 151
 152 struct SearchResult
 153 {
 154         QValueList<Entry> list;
 155         QStringList results;
 156         int count, outOf;
 157         bool common;
 158         QString text;
 159 };
 160
 161 enum SearchType { Search_Beginning, Search_FullWord, Search_Anywhere };
 162 enum DictionaryType { Edict, Kanjidict };
 163
 164 class KDE_EXPORT Index : public QObject
 165 {
 166 Q_OBJECT
 167
 168 public:
 169         Index();
 170         virtual ~Index();
 171
 172         void setDictList(const QStringList &files, const QStringList &names);
 173         void setKanjiDictList(const QStringList &files, const QStringList &names);
 174
 175         SearchResult search(QRegExp, const QString &, bool common);
 176         SearchResult searchKanji(QRegExp, const QString &, bool common);
 177         SearchResult searchPrevious(QRegExp, const QString &, SearchResult, bool common);
 178
 179         // convenience function to create suitable regexps
 180         static QRegExp createRegExp(SearchType type, const QString &text, DictionaryType dictionaryType, bool caseSensitive = false);
 181
 182 private:
 183         QPtrList<File> dictFiles;
 184         QPtrList<File> kanjiDictFiles;
 185
 186         void loadDictList(QPtrList<File> &fileList, const QStringList &dictList, const QStringList &dictNameList);
 187
 188         QStringList doSearch(File &, const QString &);
 189         SearchResult scanResults(QRegExp regexp, QStringList results, bool common);
 190         SearchResult scanKanjiResults(QRegExp regexp, QStringList results, bool common);
 191         int stringCompare(File &, int index, QCString);
 192 };
 193
 194 // lotsa helper functions
 195 KDE_EXPORT QString prettyKanjiReading(QStringList);
 196 KDE_EXPORT QString prettyMeaning(QStringList);
 197 KDE_EXPORT Entry parse(const QString &);
 198 KDE_EXPORT Entry kanjiParse(const QString &);
 199 KDE_EXPORT Dict::Entry firstEntry(Dict::SearchResult);
 200 KDE_EXPORT QString firstEntryText(Dict::SearchResult);
 201
 202 int eucStringCompare(const char *str1, const char *str2);
 203 bool isEUC(unsigned char c);
 204 }
 205
 206 #endif