kdict.h

   1 /*
   2 Project: J-Ben
   3 Author:  Paul Goins
   4 Website: http://www.vultaire.net/software/jben/
   5 License: GNU General Public License (GPL) version 2
   6          (http://www.gnu.org/licenses/old-licenses/gpl-2.0.txt)
   7
   8 File: kdict.h
   9
  10 This program is free software; you can redistribute it and/or modify
  11 it under the terms of the GNU General Public License as published by
  12 the Free Software Foundation; either version 2 of the License, or
  13 (at your option) any later version.
  14
  15 This program is distributed in the hope that it will be useful,
  16 but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 GNU General Public License for more details.
  19
  20 You should have received a copy of the GNU General Public License
  21 along with this program.  If not, see <http://www.gnu.org/licenses/>
  22 */
  23
  24 #ifndef kdict_h
  25 #define kdict_h
  26
  27 #define KD_SUCCESS 0x0
  28 #define KD_FAILURE 0x80000000
  29
  30 /* Onyomi/kunyomi/nanori/radical readings */
  31 #define KDO_READINGS       0x1
  32 /* English meanings */
  33 #define KDO_MEANINGS       0x2
  34 /* Grade, frequency, stroke count(s) */
  35 #define KDO_HIGHIMPORTANCE 0x4
  36 /* Enable dictionary display */
  37 #define KDO_DICTIONARIES   0x8
  38 /* Less important miscellanea, plus radical #'s */
  39 #define KDO_LOWIMPORTANCE 0x10
  40 /* Enable cross-referencing with vocab study list */
  41 #define KDO_VOCABCROSSREF 0x20
  42 /* KanjiCafe.com Stroke Order Diagrams (if present) */
  43 #define KDO_SOD_STATIC    0x40
  44 #define KDO_SOD_ANIM      0x80
  45 /* All options (except UNHANDLED) */
  46 #define KDO_ALL           0xFF
  47 /* All unhandled dictionary stuff */
  48 #define KDO_UNHANDLED   0x8000
  49
  50 /* Hn...: "New Japanese-English Character Dictionary" by Jack Halpern.  "Halpern" field 1 in JWPce. */
  51 #define KDD_NJECD  0x1
  52 /* Nn...: "Modern Reader's Japanese-English Character Dictionary" by Andrew Nelson.  "Nelson" field 1 in JWPce. */
  53 #define KDD_MRJECD 0x2
  54 /* Vn...: "The New Nelson Japanese-English Character Dictionary" by John Haig.  "Nelson" field 2 in JWPce. */
  55 #define KDD_NNJECD 0x4
  56 /* More "D" codes... add as needed */
  57 /* DBnnn: Japanese for Busy People, AJLT */
  58 #define KDD_JBP    0x8
  59 /* DCnnnn: The Kanji Way to Japanese Language Power, Dale Crowley */
  60 #define KDD_KWJLP  0x10
  61 /* DGnnn: Kodansha Compact Kanji Guide */
  62 #define KDD_KCKG   0x20
  63 /* DHnnnn: A Guide To Reading and Writing Japanese, Ken Hensall et al. */
  64 #define KDD_GTRWJH 0x40
  65 /* DJnnnn: Kanji in Context, Nishiguchi and Kono */
  66 #define KDD_KIC    0x80
  67 /* DKnnnn: Kanji Learners Dictionary, Jack Halpern */
  68 #define KDD_KLD    0x100
  69 /* DOnnnn: Essential Kanji, P.G. O'Neill */
  70 #define KDD_EK     0x200
  71 /* DRnnnn: 2001 Kanji, Father Joseph De Roo */
  72 #define KDD_DR     0x400
  73 /* DSnnnn: A Guide To Reading and Writing Japanese, Florence Sakade */
  74 #define KDD_GTRWJS 0x800
  75 /* DTnnn: Tuttle Kanji Cards, Alexander Kask */
  76 #define KDD_TKC    0x1000
  77 /* DFnnnn: Japanese Kanji Flashcards, White Rabbit Press */
  78 #define KDD_JKF    0x2000
  79 /* "Pn...-n...-n..." - SKIP codes used by Halpern dictionaries */
  80 #define KDD_SKIP   0x4000
  81 /* Inxnn.n: Kanji Dictionary, Spahn & Hadamitzky */
  82 #define KDD_KD     0x8000
  83 /* INnnnn: Kanji & Kana, Spahn & Hadamitzky */
  84 #define KDD_KK     0x10000
  85 /* Qnnnn.n: Four Corner code, used by various dictionaries in China and Japan */
  86 #define KDD_FC     0x20000
  87 /* MNnnnnnnn: Morohashi Daikanwajiten, index number */
  88 #define KDD_MOROI  0x40000
  89 /* MPnn.nnnn: Morohashi Daikanwajiten, volume.page number */
  90 #define KDD_MOROVP 0x80000
  91 /* Morohashi Daikanwajiten (Both indexing options) */
  92 #define KDD_MORO   KDD_MOROI | KDD_MOROVP;
  93 /* Ennnn: A Guide to Remembering Japanese Characters, Kenneth G. Henshal */
  94 #define KDD_GRJC   0x100000
  95 /* Knnnn: Gakken Kanji Dictionary ("A New Dictionary of Kanji Usage") */
  96 #define KDD_GKD    0x200000
  97 /* Lnnnn: Remembering The Kanji, James Heisig */
  98 #define KDD_RTK    0x400000
  99 /* Onnnn[A]: Japanese Names, P.G. O'Neill */
 100 #define KDD_JN     0x800000
 101 /* Show EVERYTHING!
 102    This is a lot of output, so it's probably for debug purposes only. */
 103 #define KDD_ALL    0xFFFFFFFF
 104
 105 #include "boosthm.h"
 106 #include <string>
 107 #include <iostream>
 108 #include <list>
 109 #include <map>
 110 using namespace std;
 111
 112 class KInfo {
 113 public:
 114         KInfo();
 115
 116         string literal;
 117         map<string,string> codepoint;
 118         unsigned char radical, radicalNelson;
 119         int grade;
 120         int strokeCount;
 121         list<int> misstrokes;
 122         map<string,string> variant;
 123         int freq;
 124         string radicalName;
 125         map<string,string> dictCode;
 126         map<string,string> queryCode;
 127         list< pair<string,string> > skipMisclass;
 128         list<string> pinyin, korean_r, korean_h,
 129                 onyomi, kunyomi, nanori;
 130         /* FOR NOW: I am ignoring onyomi and kunyomi r_status and on_type flags.
 131            In the KANJIDIC2 dated 2008-02-12, there was only one entry, "行く",
 132            which had any of these flags.  Not worth the effort. */
 133         map<string, list<string> > meaning;
 134 };
 135
 136 class KDict {
 137 public:
 138         static const KDict *Get();
 139         static void Destroy();
 140         ~KDict();
 141
 142         /* General purpose access functions */
 143         static wstring KInfoToHtml(const KInfo& kInfo);
 144         static wstring KInfoToHtml(const KInfo& kInfo,
 145                                                            long options, long dictionaries);
 146         const KInfo* GetEntry(const wchar_t key) const;
 147         const BoostHM<wchar_t, KInfo>* GetHashTable() const;
 148
 149         /* Other functions */
 150         bool MainDataLoaded() const;
 151 private:
 152         /* Hidden constructor */
 153         KDict();
 154
 155         /* Dictionary file loaders */
 156         int LoadKanjidic(const char* filename, const char* jisStd="jis208");
 157         int LoadKanjidic2(const char* filename);
 158         int LoadRadkfile(const char* filename);
 159         int LoadKradfile(const char* filename);
 160
 161         /* Kanjidic-specific stuff */
 162         void KanjidicToKInfo(const string& kanjidicEntry, KInfo& k,
 163                                                  const char* jisStd);
 164         void KanjidicParser(char* kanjidicRawData, const char* jisStd);
 165         static wstring ConvertKanjidicEntry(const wstring& s);
 166
 167         /* Data */
 168         static KDict *kdictSingleton;
 169         BoostHM<wchar_t, KInfo> kdictData;
 170         BoostHM<wchar_t, wstring> radkData, kradData;
 171         BoostHM<wchar_t, int> radkDataStrokes;
 172 };
 173
 174 #endif