1diff --git a/dict/main.c b/dict/main.c
2index e5573fea55a25a90f4ad79493f38ec58e3886a3b..b743e07a307618cbf522b7a770da37953dd605f2 100644
3--- a/dict/main.c
4+++ b/dict/main.c
5@@ -14,24 +14,25 @@
6 void search(char*, int);
7 int run(const char*, const char*);
8
9-int main(int argc, char** argv) {
10+int main(int argc, char** argv)
11+{
12 int opt;
13 char* txt = NULL;
14 char* db = NULL;
15
16 while ((opt = getopt(argc, argv, "t:d:h")) != -1) {
17 switch(opt) {
18- case 't':
19- txt = copy_achar(optarg);
20- break;
21- case 'd':
22- db = copy_achar(optarg);
23- break;
24- case 'h':
25- // fall through
26- default:
27- printf("Usage: %s", argv[0]);
28- goto end;
29+ case 't':
30+ txt = copy_achar(optarg);
31+ break;
32+ case 'd':
33+ db = copy_achar(optarg);
34+ break;
35+ case 'h':
36+ // fall through
37+ default:
38+ printf("Usage: %s", argv[0]);
39+ goto end;
40 }
41 }
42
43@@ -46,7 +47,8 @@
44 return r;
45 }
46
47-int run(const char *db, const char *txt) {
48+int run(const char *db, const char *txt)
49+{
50 data = new_data(db);
51 bootstrap(data);
52
53@@ -92,7 +94,8 @@ free_data(data);
54 return 0;
55 }
56
57-void search(char *sch, int len) {
58+void search(char *sch, int len)
59+{
60 char s[len+2];
61
62 sprintf(s, "%%%*s%%", len, sch);
63@@ -107,5 +110,3 @@ printw("%s", item->Line);
64 }
65 refresh();
66 }
67-
68-
69diff --git a/ext/spellfix.c b/ext/spellfix.c
70index a0c5aafd1009b6cfe79f5d46c98572f77a335706..77840a56ca6b73a38a338ffefd7bee2ef009f02c 100644
71--- a/ext/spellfix.c
72+++ b/ext/spellfix.c
73@@ -30,8 +30,8 @@ # include <stdlib.h>
74 # include <assert.h>
75 # define ALWAYS(X) 1
76 # define NEVER(X) 0
77- typedef unsigned char u8;
78- typedef unsigned short u16;
79+typedef unsigned char u8;
80+typedef unsigned short u16;
81 #endif
82 #include <ctype.h>
83
84@@ -73,99 +73,99 @@ ** The following table gives the character class for non-initial ASCII
85 ** characters.
86 */
87 static const unsigned char midClass[] = {
88- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
89- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
90- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
91- /* */ CCLASS_SPACE, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
92- /* */ CCLASS_SPACE, /* */ CCLASS_SPACE, /* */ CCLASS_OTHER,
93- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
94- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
95- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
96- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
97- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
98- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_SPACE,
99- /* ! */ CCLASS_OTHER, /* " */ CCLASS_OTHER, /* # */ CCLASS_OTHER,
100- /* $ */ CCLASS_OTHER, /* % */ CCLASS_OTHER, /* & */ CCLASS_OTHER,
101- /* ' */ CCLASS_SILENT, /* ( */ CCLASS_OTHER, /* ) */ CCLASS_OTHER,
102- /* * */ CCLASS_OTHER, /* + */ CCLASS_OTHER, /* , */ CCLASS_OTHER,
103- /* - */ CCLASS_OTHER, /* . */ CCLASS_OTHER, /* / */ CCLASS_OTHER,
104- /* 0 */ CCLASS_DIGIT, /* 1 */ CCLASS_DIGIT, /* 2 */ CCLASS_DIGIT,
105- /* 3 */ CCLASS_DIGIT, /* 4 */ CCLASS_DIGIT, /* 5 */ CCLASS_DIGIT,
106- /* 6 */ CCLASS_DIGIT, /* 7 */ CCLASS_DIGIT, /* 8 */ CCLASS_DIGIT,
107- /* 9 */ CCLASS_DIGIT, /* : */ CCLASS_OTHER, /* ; */ CCLASS_OTHER,
108- /* < */ CCLASS_OTHER, /* = */ CCLASS_OTHER, /* > */ CCLASS_OTHER,
109- /* ? */ CCLASS_OTHER, /* @ */ CCLASS_OTHER, /* A */ CCLASS_VOWEL,
110- /* B */ CCLASS_B, /* C */ CCLASS_C, /* D */ CCLASS_D,
111- /* E */ CCLASS_VOWEL, /* F */ CCLASS_B, /* G */ CCLASS_C,
112- /* H */ CCLASS_SILENT, /* I */ CCLASS_VOWEL, /* J */ CCLASS_C,
113- /* K */ CCLASS_C, /* L */ CCLASS_L, /* M */ CCLASS_M,
114- /* N */ CCLASS_M, /* O */ CCLASS_VOWEL, /* P */ CCLASS_B,
115- /* Q */ CCLASS_C, /* R */ CCLASS_R, /* S */ CCLASS_C,
116- /* T */ CCLASS_D, /* U */ CCLASS_VOWEL, /* V */ CCLASS_B,
117- /* W */ CCLASS_B, /* X */ CCLASS_C, /* Y */ CCLASS_VOWEL,
118- /* Z */ CCLASS_C, /* [ */ CCLASS_OTHER, /* \ */ CCLASS_OTHER,
119- /* ] */ CCLASS_OTHER, /* ^ */ CCLASS_OTHER, /* _ */ CCLASS_OTHER,
120- /* ` */ CCLASS_OTHER, /* a */ CCLASS_VOWEL, /* b */ CCLASS_B,
121- /* c */ CCLASS_C, /* d */ CCLASS_D, /* e */ CCLASS_VOWEL,
122- /* f */ CCLASS_B, /* g */ CCLASS_C, /* h */ CCLASS_SILENT,
123- /* i */ CCLASS_VOWEL, /* j */ CCLASS_C, /* k */ CCLASS_C,
124- /* l */ CCLASS_L, /* m */ CCLASS_M, /* n */ CCLASS_M,
125- /* o */ CCLASS_VOWEL, /* p */ CCLASS_B, /* q */ CCLASS_C,
126- /* r */ CCLASS_R, /* s */ CCLASS_C, /* t */ CCLASS_D,
127- /* u */ CCLASS_VOWEL, /* v */ CCLASS_B, /* w */ CCLASS_B,
128- /* x */ CCLASS_C, /* y */ CCLASS_VOWEL, /* z */ CCLASS_C,
129- /* { */ CCLASS_OTHER, /* | */ CCLASS_OTHER, /* } */ CCLASS_OTHER,
130- /* ~ */ CCLASS_OTHER, /* */ CCLASS_OTHER,
131+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
132+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
133+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
134+ /* */ CCLASS_SPACE, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
135+ /* */ CCLASS_SPACE, /* */ CCLASS_SPACE, /* */ CCLASS_OTHER,
136+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
137+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
138+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
139+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
140+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
141+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_SPACE,
142+ /* ! */ CCLASS_OTHER, /* " */ CCLASS_OTHER, /* # */ CCLASS_OTHER,
143+ /* $ */ CCLASS_OTHER, /* % */ CCLASS_OTHER, /* & */ CCLASS_OTHER,
144+ /* ' */ CCLASS_SILENT, /* ( */ CCLASS_OTHER, /* ) */ CCLASS_OTHER,
145+ /* * */ CCLASS_OTHER, /* + */ CCLASS_OTHER, /* , */ CCLASS_OTHER,
146+ /* - */ CCLASS_OTHER, /* . */ CCLASS_OTHER, /* / */ CCLASS_OTHER,
147+ /* 0 */ CCLASS_DIGIT, /* 1 */ CCLASS_DIGIT, /* 2 */ CCLASS_DIGIT,
148+ /* 3 */ CCLASS_DIGIT, /* 4 */ CCLASS_DIGIT, /* 5 */ CCLASS_DIGIT,
149+ /* 6 */ CCLASS_DIGIT, /* 7 */ CCLASS_DIGIT, /* 8 */ CCLASS_DIGIT,
150+ /* 9 */ CCLASS_DIGIT, /* : */ CCLASS_OTHER, /* ; */ CCLASS_OTHER,
151+ /* < */ CCLASS_OTHER, /* = */ CCLASS_OTHER, /* > */ CCLASS_OTHER,
152+ /* ? */ CCLASS_OTHER, /* @ */ CCLASS_OTHER, /* A */ CCLASS_VOWEL,
153+ /* B */ CCLASS_B, /* C */ CCLASS_C, /* D */ CCLASS_D,
154+ /* E */ CCLASS_VOWEL, /* F */ CCLASS_B, /* G */ CCLASS_C,
155+ /* H */ CCLASS_SILENT, /* I */ CCLASS_VOWEL, /* J */ CCLASS_C,
156+ /* K */ CCLASS_C, /* L */ CCLASS_L, /* M */ CCLASS_M,
157+ /* N */ CCLASS_M, /* O */ CCLASS_VOWEL, /* P */ CCLASS_B,
158+ /* Q */ CCLASS_C, /* R */ CCLASS_R, /* S */ CCLASS_C,
159+ /* T */ CCLASS_D, /* U */ CCLASS_VOWEL, /* V */ CCLASS_B,
160+ /* W */ CCLASS_B, /* X */ CCLASS_C, /* Y */ CCLASS_VOWEL,
161+ /* Z */ CCLASS_C, /* [ */ CCLASS_OTHER, /* \ */ CCLASS_OTHER,
162+ /* ] */ CCLASS_OTHER, /* ^ */ CCLASS_OTHER, /* _ */ CCLASS_OTHER,
163+ /* ` */ CCLASS_OTHER, /* a */ CCLASS_VOWEL, /* b */ CCLASS_B,
164+ /* c */ CCLASS_C, /* d */ CCLASS_D, /* e */ CCLASS_VOWEL,
165+ /* f */ CCLASS_B, /* g */ CCLASS_C, /* h */ CCLASS_SILENT,
166+ /* i */ CCLASS_VOWEL, /* j */ CCLASS_C, /* k */ CCLASS_C,
167+ /* l */ CCLASS_L, /* m */ CCLASS_M, /* n */ CCLASS_M,
168+ /* o */ CCLASS_VOWEL, /* p */ CCLASS_B, /* q */ CCLASS_C,
169+ /* r */ CCLASS_R, /* s */ CCLASS_C, /* t */ CCLASS_D,
170+ /* u */ CCLASS_VOWEL, /* v */ CCLASS_B, /* w */ CCLASS_B,
171+ /* x */ CCLASS_C, /* y */ CCLASS_VOWEL, /* z */ CCLASS_C,
172+ /* { */ CCLASS_OTHER, /* | */ CCLASS_OTHER, /* } */ CCLASS_OTHER,
173+ /* ~ */ CCLASS_OTHER, /* */ CCLASS_OTHER,
174 };
175-/*
176+/*
177 ** This tables gives the character class for ASCII characters that form the
178 ** initial character of a word. The only difference from midClass is with
179 ** the letters H, W, and Y.
180 */
181 static const unsigned char initClass[] = {
182- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
183- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
184- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
185- /* */ CCLASS_SPACE, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
186- /* */ CCLASS_SPACE, /* */ CCLASS_SPACE, /* */ CCLASS_OTHER,
187- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
188- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
189- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
190- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
191- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
192- /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_SPACE,
193- /* ! */ CCLASS_OTHER, /* " */ CCLASS_OTHER, /* # */ CCLASS_OTHER,
194- /* $ */ CCLASS_OTHER, /* % */ CCLASS_OTHER, /* & */ CCLASS_OTHER,
195- /* ' */ CCLASS_OTHER, /* ( */ CCLASS_OTHER, /* ) */ CCLASS_OTHER,
196- /* * */ CCLASS_OTHER, /* + */ CCLASS_OTHER, /* , */ CCLASS_OTHER,
197- /* - */ CCLASS_OTHER, /* . */ CCLASS_OTHER, /* / */ CCLASS_OTHER,
198- /* 0 */ CCLASS_DIGIT, /* 1 */ CCLASS_DIGIT, /* 2 */ CCLASS_DIGIT,
199- /* 3 */ CCLASS_DIGIT, /* 4 */ CCLASS_DIGIT, /* 5 */ CCLASS_DIGIT,
200- /* 6 */ CCLASS_DIGIT, /* 7 */ CCLASS_DIGIT, /* 8 */ CCLASS_DIGIT,
201- /* 9 */ CCLASS_DIGIT, /* : */ CCLASS_OTHER, /* ; */ CCLASS_OTHER,
202- /* < */ CCLASS_OTHER, /* = */ CCLASS_OTHER, /* > */ CCLASS_OTHER,
203- /* ? */ CCLASS_OTHER, /* @ */ CCLASS_OTHER, /* A */ CCLASS_VOWEL,
204- /* B */ CCLASS_B, /* C */ CCLASS_C, /* D */ CCLASS_D,
205- /* E */ CCLASS_VOWEL, /* F */ CCLASS_B, /* G */ CCLASS_C,
206- /* H */ CCLASS_SILENT, /* I */ CCLASS_VOWEL, /* J */ CCLASS_C,
207- /* K */ CCLASS_C, /* L */ CCLASS_L, /* M */ CCLASS_M,
208- /* N */ CCLASS_M, /* O */ CCLASS_VOWEL, /* P */ CCLASS_B,
209- /* Q */ CCLASS_C, /* R */ CCLASS_R, /* S */ CCLASS_C,
210- /* T */ CCLASS_D, /* U */ CCLASS_VOWEL, /* V */ CCLASS_B,
211- /* W */ CCLASS_B, /* X */ CCLASS_C, /* Y */ CCLASS_Y,
212- /* Z */ CCLASS_C, /* [ */ CCLASS_OTHER, /* \ */ CCLASS_OTHER,
213- /* ] */ CCLASS_OTHER, /* ^ */ CCLASS_OTHER, /* _ */ CCLASS_OTHER,
214- /* ` */ CCLASS_OTHER, /* a */ CCLASS_VOWEL, /* b */ CCLASS_B,
215- /* c */ CCLASS_C, /* d */ CCLASS_D, /* e */ CCLASS_VOWEL,
216- /* f */ CCLASS_B, /* g */ CCLASS_C, /* h */ CCLASS_SILENT,
217- /* i */ CCLASS_VOWEL, /* j */ CCLASS_C, /* k */ CCLASS_C,
218- /* l */ CCLASS_L, /* m */ CCLASS_M, /* n */ CCLASS_M,
219- /* o */ CCLASS_VOWEL, /* p */ CCLASS_B, /* q */ CCLASS_C,
220- /* r */ CCLASS_R, /* s */ CCLASS_C, /* t */ CCLASS_D,
221- /* u */ CCLASS_VOWEL, /* v */ CCLASS_B, /* w */ CCLASS_B,
222- /* x */ CCLASS_C, /* y */ CCLASS_Y, /* z */ CCLASS_C,
223- /* { */ CCLASS_OTHER, /* | */ CCLASS_OTHER, /* } */ CCLASS_OTHER,
224- /* ~ */ CCLASS_OTHER, /* */ CCLASS_OTHER,
225+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
226+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
227+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
228+ /* */ CCLASS_SPACE, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
229+ /* */ CCLASS_SPACE, /* */ CCLASS_SPACE, /* */ CCLASS_OTHER,
230+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
231+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
232+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
233+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
234+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_OTHER,
235+ /* */ CCLASS_OTHER, /* */ CCLASS_OTHER, /* */ CCLASS_SPACE,
236+ /* ! */ CCLASS_OTHER, /* " */ CCLASS_OTHER, /* # */ CCLASS_OTHER,
237+ /* $ */ CCLASS_OTHER, /* % */ CCLASS_OTHER, /* & */ CCLASS_OTHER,
238+ /* ' */ CCLASS_OTHER, /* ( */ CCLASS_OTHER, /* ) */ CCLASS_OTHER,
239+ /* * */ CCLASS_OTHER, /* + */ CCLASS_OTHER, /* , */ CCLASS_OTHER,
240+ /* - */ CCLASS_OTHER, /* . */ CCLASS_OTHER, /* / */ CCLASS_OTHER,
241+ /* 0 */ CCLASS_DIGIT, /* 1 */ CCLASS_DIGIT, /* 2 */ CCLASS_DIGIT,
242+ /* 3 */ CCLASS_DIGIT, /* 4 */ CCLASS_DIGIT, /* 5 */ CCLASS_DIGIT,
243+ /* 6 */ CCLASS_DIGIT, /* 7 */ CCLASS_DIGIT, /* 8 */ CCLASS_DIGIT,
244+ /* 9 */ CCLASS_DIGIT, /* : */ CCLASS_OTHER, /* ; */ CCLASS_OTHER,
245+ /* < */ CCLASS_OTHER, /* = */ CCLASS_OTHER, /* > */ CCLASS_OTHER,
246+ /* ? */ CCLASS_OTHER, /* @ */ CCLASS_OTHER, /* A */ CCLASS_VOWEL,
247+ /* B */ CCLASS_B, /* C */ CCLASS_C, /* D */ CCLASS_D,
248+ /* E */ CCLASS_VOWEL, /* F */ CCLASS_B, /* G */ CCLASS_C,
249+ /* H */ CCLASS_SILENT, /* I */ CCLASS_VOWEL, /* J */ CCLASS_C,
250+ /* K */ CCLASS_C, /* L */ CCLASS_L, /* M */ CCLASS_M,
251+ /* N */ CCLASS_M, /* O */ CCLASS_VOWEL, /* P */ CCLASS_B,
252+ /* Q */ CCLASS_C, /* R */ CCLASS_R, /* S */ CCLASS_C,
253+ /* T */ CCLASS_D, /* U */ CCLASS_VOWEL, /* V */ CCLASS_B,
254+ /* W */ CCLASS_B, /* X */ CCLASS_C, /* Y */ CCLASS_Y,
255+ /* Z */ CCLASS_C, /* [ */ CCLASS_OTHER, /* \ */ CCLASS_OTHER,
256+ /* ] */ CCLASS_OTHER, /* ^ */ CCLASS_OTHER, /* _ */ CCLASS_OTHER,
257+ /* ` */ CCLASS_OTHER, /* a */ CCLASS_VOWEL, /* b */ CCLASS_B,
258+ /* c */ CCLASS_C, /* d */ CCLASS_D, /* e */ CCLASS_VOWEL,
259+ /* f */ CCLASS_B, /* g */ CCLASS_C, /* h */ CCLASS_SILENT,
260+ /* i */ CCLASS_VOWEL, /* j */ CCLASS_C, /* k */ CCLASS_C,
261+ /* l */ CCLASS_L, /* m */ CCLASS_M, /* n */ CCLASS_M,
262+ /* o */ CCLASS_VOWEL, /* p */ CCLASS_B, /* q */ CCLASS_C,
263+ /* r */ CCLASS_R, /* s */ CCLASS_C, /* t */ CCLASS_D,
264+ /* u */ CCLASS_VOWEL, /* v */ CCLASS_B, /* w */ CCLASS_B,
265+ /* x */ CCLASS_C, /* y */ CCLASS_Y, /* z */ CCLASS_C,
266+ /* { */ CCLASS_OTHER, /* | */ CCLASS_OTHER, /* } */ CCLASS_OTHER,
267+ /* ~ */ CCLASS_OTHER, /* */ CCLASS_OTHER,
268 };
269
270 /*
271@@ -189,54 +189,58 @@ ** * Omit K in KN or G in GN at the beginning of a word
272 **
273 ** Space to hold the result is obtained from sqlite3_malloc()
274 **
275-** Return NULL if memory allocation fails.
276+** Return NULL if memory allocation fails.
277 */
278-static unsigned char *phoneticHash(const unsigned char *zIn, int nIn){
279- unsigned char *zOut = sqlite3_malloc64( nIn + 1 );
280- int i;
281- int nOut = 0;
282- char cPrev = 0x77;
283- char cPrevX = 0x77;
284- const unsigned char *aClass = initClass;
285+static unsigned char *phoneticHash(const unsigned char *zIn, int nIn)
286+{
287+ unsigned char *zOut = sqlite3_malloc64( nIn + 1 );
288+ int i;
289+ int nOut = 0;
290+ char cPrev = 0x77;
291+ char cPrevX = 0x77;
292+ const unsigned char *aClass = initClass;
293
294- if( zOut==0 ) return 0;
295- if( nIn>2 ){
296- switch( zIn[0] ){
297- case 'g':
298- case 'k': {
299- if( zIn[1]=='n' ){ zIn++; nIn--; }
300- break;
301- }
302- }
303- }
304- for(i=0; i<nIn; i++){
305- unsigned char c = zIn[i];
306- if( i+1<nIn ){
307- if( c=='w' && zIn[i+1]=='r' ) continue;
308- if( c=='d' && (zIn[i+1]=='j' || zIn[i+1]=='g') ) continue;
309- if( i+2<nIn ){
310- if( c=='t' && zIn[i+1]=='c' && zIn[i+2]=='h' ) continue;
311- }
312- }
313- c = aClass[c&0x7f];
314- if( c==CCLASS_SPACE ) continue;
315- if( c==CCLASS_OTHER && cPrev!=CCLASS_DIGIT ) continue;
316- aClass = midClass;
317- if( c==CCLASS_VOWEL && (cPrevX==CCLASS_R || cPrevX==CCLASS_L) ){
318- continue; /* No vowels beside L or R */
319+ if( zOut==0 ) return 0;
320+ if( nIn>2 ) {
321+ switch( zIn[0] ) {
322+ case 'g':
323+ case 'k': {
324+ if( zIn[1]=='n' ) {
325+ zIn++;
326+ nIn--;
327+ }
328+ break;
329+ }
330+ }
331 }
332- if( (c==CCLASS_R || c==CCLASS_L) && cPrevX==CCLASS_VOWEL ){
333- nOut--; /* No vowels beside L or R */
334+ for(i=0; i<nIn; i++) {
335+ unsigned char c = zIn[i];
336+ if( i+1<nIn ) {
337+ if( c=='w' && zIn[i+1]=='r' ) continue;
338+ if( c=='d' && (zIn[i+1]=='j' || zIn[i+1]=='g') ) continue;
339+ if( i+2<nIn ) {
340+ if( c=='t' && zIn[i+1]=='c' && zIn[i+2]=='h' ) continue;
341+ }
342+ }
343+ c = aClass[c&0x7f];
344+ if( c==CCLASS_SPACE ) continue;
345+ if( c==CCLASS_OTHER && cPrev!=CCLASS_DIGIT ) continue;
346+ aClass = midClass;
347+ if( c==CCLASS_VOWEL && (cPrevX==CCLASS_R || cPrevX==CCLASS_L) ) {
348+ continue; /* No vowels beside L or R */
349+ }
350+ if( (c==CCLASS_R || c==CCLASS_L) && cPrevX==CCLASS_VOWEL ) {
351+ nOut--; /* No vowels beside L or R */
352+ }
353+ cPrev = c;
354+ if( c==CCLASS_SILENT ) continue;
355+ cPrevX = c;
356+ c = className[c];
357+ assert( nOut>=0 );
358+ if( nOut==0 || c!=zOut[nOut-1] ) zOut[nOut++] = c;
359 }
360- cPrev = c;
361- if( c==CCLASS_SILENT ) continue;
362- cPrevX = c;
363- c = className[c];
364- assert( nOut>=0 );
365- if( nOut==0 || c!=zOut[nOut-1] ) zOut[nOut++] = c;
366- }
367- zOut[nOut] = 0;
368- return zOut;
369+ zOut[nOut] = 0;
370+ return zOut;
371 }
372
373 /*
374@@ -244,29 +248,31 @@ ** This is an SQL function wrapper around phoneticHash(). See
375 ** the description of phoneticHash() for additional information.
376 */
377 static void phoneticHashSqlFunc(
378- sqlite3_context *context,
379- int argc,
380- sqlite3_value **argv
381-){
382- const unsigned char *zIn;
383- unsigned char *zOut;
384+ sqlite3_context *context,
385+ int argc,
386+ sqlite3_value **argv
387+)
388+{
389+ const unsigned char *zIn;
390+ unsigned char *zOut;
391
392- zIn = sqlite3_value_text(argv[0]);
393- if( zIn==0 ) return;
394- zOut = phoneticHash(zIn, sqlite3_value_bytes(argv[0]));
395- if( zOut==0 ){
396- sqlite3_result_error_nomem(context);
397- }else{
398- sqlite3_result_text(context, (char*)zOut, -1, sqlite3_free);
399- }
400+ zIn = sqlite3_value_text(argv[0]);
401+ if( zIn==0 ) return;
402+ zOut = phoneticHash(zIn, sqlite3_value_bytes(argv[0]));
403+ if( zOut==0 ) {
404+ sqlite3_result_error_nomem(context);
405+ } else {
406+ sqlite3_result_text(context, (char*)zOut, -1, sqlite3_free);
407+ }
408 }
409
410 /*
411 ** Return the character class number for a character given its
412 ** context.
413 */
414-static char characterClass(char cPrev, char c){
415- return cPrev==0 ? initClass[c&0x7f] : midClass[c&0x7f];
416+static char characterClass(char cPrev, char c)
417+{
418+ return cPrev==0 ? initClass[c&0x7f] : midClass[c&0x7f];
419 }
420
421 /*
422@@ -274,34 +280,35 @@ ** Return the cost of inserting or deleting character c immediately
423 ** following character cPrev. If cPrev==0, that means c is the first
424 ** character of the word.
425 */
426-static int insertOrDeleteCost(char cPrev, char c, char cNext){
427- char classC = characterClass(cPrev, c);
428- char classCprev;
429+static int insertOrDeleteCost(char cPrev, char c, char cNext)
430+{
431+ char classC = characterClass(cPrev, c);
432+ char classCprev;
433
434- if( classC==CCLASS_SILENT ){
435- /* Insert or delete "silent" characters such as H or W */
436- return 1;
437- }
438- if( cPrev==c ){
439- /* Repeated characters, or miss a repeat */
440- return 10;
441- }
442- if( classC==CCLASS_VOWEL && (cPrev=='r' || cNext=='r') ){
443- return 20; /* Insert a vowel before or after 'r' */
444- }
445- classCprev = characterClass(cPrev, cPrev);
446- if( classC==classCprev ){
447- if( classC==CCLASS_VOWEL ){
448- /* Remove or add a new vowel to a vowel cluster */
449- return 15;
450- }else{
451- /* Remove or add a consonant not in the same class */
452- return 50;
453+ if( classC==CCLASS_SILENT ) {
454+ /* Insert or delete "silent" characters such as H or W */
455+ return 1;
456 }
457- }
458+ if( cPrev==c ) {
459+ /* Repeated characters, or miss a repeat */
460+ return 10;
461+ }
462+ if( classC==CCLASS_VOWEL && (cPrev=='r' || cNext=='r') ) {
463+ return 20; /* Insert a vowel before or after 'r' */
464+ }
465+ classCprev = characterClass(cPrev, cPrev);
466+ if( classC==classCprev ) {
467+ if( classC==CCLASS_VOWEL ) {
468+ /* Remove or add a new vowel to a vowel cluster */
469+ return 15;
470+ } else {
471+ /* Remove or add a consonant not in the same class */
472+ return 50;
473+ }
474+ }
475
476- /* any other character insertion or deletion */
477- return 100;
478+ /* any other character insertion or deletion */
479+ return 100;
480 }
481
482 /*
483@@ -315,29 +322,30 @@ ** Return the cost of substituting cTo in place of cFrom assuming
484 ** the previous character is cPrev. If cPrev==0 then cTo is the first
485 ** character of the word.
486 */
487-static int substituteCost(char cPrev, char cFrom, char cTo){
488- char classFrom, classTo;
489- if( cFrom==cTo ){
490- /* Exact match */
491- return 0;
492- }
493- if( cFrom==(cTo^0x20) && ((cTo>='A' && cTo<='Z') || (cTo>='a' && cTo<='z')) ){
494- /* differ only in case */
495- return 0;
496- }
497- classFrom = characterClass(cPrev, cFrom);
498- classTo = characterClass(cPrev, cTo);
499- if( classFrom==classTo ){
500- /* Same character class */
501- return 40;
502- }
503- if( classFrom>=CCLASS_B && classFrom<=CCLASS_Y
504- && classTo>=CCLASS_B && classTo<=CCLASS_Y ){
505- /* Convert from one consonant to another, but in a different class */
506- return 75;
507- }
508- /* Any other subsitution */
509- return 100;
510+static int substituteCost(char cPrev, char cFrom, char cTo)
511+{
512+ char classFrom, classTo;
513+ if( cFrom==cTo ) {
514+ /* Exact match */
515+ return 0;
516+ }
517+ if( cFrom==(cTo^0x20) && ((cTo>='A' && cTo<='Z') || (cTo>='a' && cTo<='z')) ) {
518+ /* differ only in case */
519+ return 0;
520+ }
521+ classFrom = characterClass(cPrev, cFrom);
522+ classTo = characterClass(cPrev, cTo);
523+ if( classFrom==classTo ) {
524+ /* Same character class */
525+ return 40;
526+ }
527+ if( classFrom>=CCLASS_B && classFrom<=CCLASS_Y
528+ && classTo>=CCLASS_B && classTo<=CCLASS_Y ) {
529+ /* Convert from one consonant to another, but in a different class */
530+ return 75;
531+ }
532+ /* Any other subsitution */
533+ return 100;
534 }
535
536 /*
537@@ -351,7 +359,7 @@ **
538 ** Negative values indicate an error:
539 ** -1 One of the inputs is NULL
540 ** -2 Non-ASCII characters on input
541-** -3 Unable to allocate memory
542+** -3 Unable to allocate memory
543 **
544 ** If pnMatch is not NULL, then *pnMatch is set to the number of bytes
545 ** of zB that matched the pattern in zA. If zA does not end with a '*',
546@@ -359,152 +367,158 @@ ** then this value is always the number of bytes in zB (i.e. strlen(zB)).
547 ** If zA does end in a '*', then it is the number of bytes in the prefix
548 ** of zB that was deemed to match zA.
549 */
550-static int editdist1(const char *zA, const char *zB, int *pnMatch){
551- int nA, nB; /* Number of characters in zA[] and zB[] */
552- int xA, xB; /* Loop counters for zA[] and zB[] */
553- char cA = 0, cB; /* Current character of zA and zB */
554- char cAprev, cBprev; /* Previous character of zA and zB */
555- char cAnext, cBnext; /* Next character in zA and zB */
556- int d; /* North-west cost value */
557- int dc = 0; /* North-west character value */
558- int res; /* Final result */
559- int *m; /* The cost matrix */
560- char *cx; /* Corresponding character values */
561- int *toFree = 0; /* Malloced space */
562- int nMatch = 0;
563- int mStack[60+15]; /* Stack space to use if not too much is needed */
564+static int editdist1(const char *zA, const char *zB, int *pnMatch)
565+{
566+ int nA, nB; /* Number of characters in zA[] and zB[] */
567+ int xA, xB; /* Loop counters for zA[] and zB[] */
568+ char cA = 0, cB; /* Current character of zA and zB */
569+ char cAprev, cBprev; /* Previous character of zA and zB */
570+ char cAnext, cBnext; /* Next character in zA and zB */
571+ int d; /* North-west cost value */
572+ int dc = 0; /* North-west character value */
573+ int res; /* Final result */
574+ int *m; /* The cost matrix */
575+ char *cx; /* Corresponding character values */
576+ int *toFree = 0; /* Malloced space */
577+ int nMatch = 0;
578+ int mStack[60+15]; /* Stack space to use if not too much is needed */
579
580- /* Early out if either input is NULL */
581- if( zA==0 || zB==0 ) return -1;
582+ /* Early out if either input is NULL */
583+ if( zA==0 || zB==0 ) return -1;
584
585- /* Skip any common prefix */
586- while( zA[0] && zA[0]==zB[0] ){ dc = zA[0]; zA++; zB++; nMatch++; }
587- if( pnMatch ) *pnMatch = nMatch;
588- if( zA[0]==0 && zB[0]==0 ) return 0;
589+ /* Skip any common prefix */
590+ while( zA[0] && zA[0]==zB[0] ) {
591+ dc = zA[0];
592+ zA++;
593+ zB++;
594+ nMatch++;
595+ }
596+ if( pnMatch ) *pnMatch = nMatch;
597+ if( zA[0]==0 && zB[0]==0 ) return 0;
598
599 #if 0
600- printf("A=\"%s\" B=\"%s\" dc=%c\n", zA, zB, dc?dc:' ');
601+ printf("A=\"%s\" B=\"%s\" dc=%c\n", zA, zB, dc?dc:' ');
602 #endif
603
604- /* Verify input strings and measure their lengths */
605- for(nA=0; zA[nA]; nA++){
606- if( zA[nA]&0x80 ) return -2;
607- }
608- for(nB=0; zB[nB]; nB++){
609- if( zB[nB]&0x80 ) return -2;
610- }
611+ /* Verify input strings and measure their lengths */
612+ for(nA=0; zA[nA]; nA++) {
613+ if( zA[nA]&0x80 ) return -2;
614+ }
615+ for(nB=0; zB[nB]; nB++) {
616+ if( zB[nB]&0x80 ) return -2;
617+ }
618
619- /* Special processing if either string is empty */
620- if( nA==0 ){
621- cBprev = (char)dc;
622- for(xB=res=0; (cB = zB[xB])!=0; xB++){
623- res += insertOrDeleteCost(cBprev, cB, zB[xB+1])/FINAL_INS_COST_DIV;
624- cBprev = cB;
625+ /* Special processing if either string is empty */
626+ if( nA==0 ) {
627+ cBprev = (char)dc;
628+ for(xB=res=0; (cB = zB[xB])!=0; xB++) {
629+ res += insertOrDeleteCost(cBprev, cB, zB[xB+1])/FINAL_INS_COST_DIV;
630+ cBprev = cB;
631+ }
632+ return res;
633 }
634- return res;
635- }
636- if( nB==0 ){
637- cAprev = (char)dc;
638- for(xA=res=0; (cA = zA[xA])!=0; xA++){
639- res += insertOrDeleteCost(cAprev, cA, zA[xA+1]);
640- cAprev = cA;
641+ if( nB==0 ) {
642+ cAprev = (char)dc;
643+ for(xA=res=0; (cA = zA[xA])!=0; xA++) {
644+ res += insertOrDeleteCost(cAprev, cA, zA[xA+1]);
645+ cAprev = cA;
646+ }
647+ return res;
648 }
649- return res;
650- }
651
652- /* A is a prefix of B */
653- if( zA[0]=='*' && zA[1]==0 ) return 0;
654+ /* A is a prefix of B */
655+ if( zA[0]=='*' && zA[1]==0 ) return 0;
656
657- /* Allocate and initialize the Wagner matrix */
658- if( nB<(sizeof(mStack)*4)/(sizeof(mStack[0])*5) ){
659- m = mStack;
660- }else{
661- m = toFree = sqlite3_malloc64( (nB+1)*5*sizeof(m[0])/4 );
662- if( m==0 ) return -3;
663- }
664- cx = (char*)&m[nB+1];
665+ /* Allocate and initialize the Wagner matrix */
666+ if( nB<(sizeof(mStack)*4)/(sizeof(mStack[0])*5) ) {
667+ m = mStack;
668+ } else {
669+ m = toFree = sqlite3_malloc64( (nB+1)*5*sizeof(m[0])/4 );
670+ if( m==0 ) return -3;
671+ }
672+ cx = (char*)&m[nB+1];
673
674- /* Compute the Wagner edit distance */
675- m[0] = 0;
676- cx[0] = (char)dc;
677- cBprev = (char)dc;
678- for(xB=1; xB<=nB; xB++){
679- cBnext = zB[xB];
680- cB = zB[xB-1];
681- cx[xB] = cB;
682- m[xB] = m[xB-1] + insertOrDeleteCost(cBprev, cB, cBnext);
683- cBprev = cB;
684- }
685- cAprev = (char)dc;
686- for(xA=1; xA<=nA; xA++){
687- int lastA = (xA==nA);
688- cA = zA[xA-1];
689- cAnext = zA[xA];
690- if( cA=='*' && lastA ) break;
691- d = m[0];
692- dc = cx[0];
693- m[0] = d + insertOrDeleteCost(cAprev, cA, cAnext);
694- cBprev = 0;
695- for(xB=1; xB<=nB; xB++){
696- int totalCost, insCost, delCost, subCost, ncx;
697- cB = zB[xB-1];
698- cBnext = zB[xB];
699+ /* Compute the Wagner edit distance */
700+ m[0] = 0;
701+ cx[0] = (char)dc;
702+ cBprev = (char)dc;
703+ for(xB=1; xB<=nB; xB++) {
704+ cBnext = zB[xB];
705+ cB = zB[xB-1];
706+ cx[xB] = cB;
707+ m[xB] = m[xB-1] + insertOrDeleteCost(cBprev, cB, cBnext);
708+ cBprev = cB;
709+ }
710+ cAprev = (char)dc;
711+ for(xA=1; xA<=nA; xA++) {
712+ int lastA = (xA==nA);
713+ cA = zA[xA-1];
714+ cAnext = zA[xA];
715+ if( cA=='*' && lastA ) break;
716+ d = m[0];
717+ dc = cx[0];
718+ m[0] = d + insertOrDeleteCost(cAprev, cA, cAnext);
719+ cBprev = 0;
720+ for(xB=1; xB<=nB; xB++) {
721+ int totalCost, insCost, delCost, subCost, ncx;
722+ cB = zB[xB-1];
723+ cBnext = zB[xB];
724
725- /* Cost to insert cB */
726- insCost = insertOrDeleteCost(cx[xB-1], cB, cBnext);
727- if( lastA ) insCost /= FINAL_INS_COST_DIV;
728+ /* Cost to insert cB */
729+ insCost = insertOrDeleteCost(cx[xB-1], cB, cBnext);
730+ if( lastA ) insCost /= FINAL_INS_COST_DIV;
731
732- /* Cost to delete cA */
733- delCost = insertOrDeleteCost(cx[xB], cA, cBnext);
734+ /* Cost to delete cA */
735+ delCost = insertOrDeleteCost(cx[xB], cA, cBnext);
736
737- /* Cost to substitute cA->cB */
738- subCost = substituteCost(cx[xB-1], cA, cB);
739+ /* Cost to substitute cA->cB */
740+ subCost = substituteCost(cx[xB-1], cA, cB);
741
742- /* Best cost */
743- totalCost = insCost + m[xB-1];
744- ncx = cB;
745- if( (delCost + m[xB])<totalCost ){
746- totalCost = delCost + m[xB];
747- ncx = cA;
748- }
749- if( (subCost + d)<totalCost ){
750- totalCost = subCost + d;
751- }
752+ /* Best cost */
753+ totalCost = insCost + m[xB-1];
754+ ncx = cB;
755+ if( (delCost + m[xB])<totalCost ) {
756+ totalCost = delCost + m[xB];
757+ ncx = cA;
758+ }
759+ if( (subCost + d)<totalCost ) {
760+ totalCost = subCost + d;
761+ }
762
763 #if 0
764- printf("%d,%d d=%4d u=%4d r=%4d dc=%c cA=%c cB=%c"
765- " ins=%4d del=%4d sub=%4d t=%4d ncx=%c\n",
766- xA, xB, d, m[xB], m[xB-1], dc?dc:' ', cA, cB,
767- insCost, delCost, subCost, totalCost, ncx?ncx:' ');
768+ printf("%d,%d d=%4d u=%4d r=%4d dc=%c cA=%c cB=%c"
769+ " ins=%4d del=%4d sub=%4d t=%4d ncx=%c\n",
770+ xA, xB, d, m[xB], m[xB-1], dc?dc:' ', cA, cB,
771+ insCost, delCost, subCost, totalCost, ncx?ncx:' ');
772 #endif
773
774- /* Update the matrix */
775- d = m[xB];
776- dc = cx[xB];
777- m[xB] = totalCost;
778- cx[xB] = (char)ncx;
779- cBprev = cB;
780+ /* Update the matrix */
781+ d = m[xB];
782+ dc = cx[xB];
783+ m[xB] = totalCost;
784+ cx[xB] = (char)ncx;
785+ cBprev = cB;
786+ }
787+ cAprev = cA;
788 }
789- cAprev = cA;
790- }
791
792- /* Free the wagner matrix and return the result */
793- if( cA=='*' ){
794- res = m[1];
795- for(xB=1; xB<=nB; xB++){
796- if( m[xB]<res ){
797- res = m[xB];
798- if( pnMatch ) *pnMatch = xB+nMatch;
799- }
800+ /* Free the wagner matrix and return the result */
801+ if( cA=='*' ) {
802+ res = m[1];
803+ for(xB=1; xB<=nB; xB++) {
804+ if( m[xB]<res ) {
805+ res = m[xB];
806+ if( pnMatch ) *pnMatch = xB+nMatch;
807+ }
808+ }
809+ } else {
810+ res = m[nB];
811+ /* In the current implementation, pnMatch is always NULL if zA does
812+ ** not end in "*" */
813+ assert( pnMatch==0 );
814 }
815- }else{
816- res = m[nB];
817- /* In the current implementation, pnMatch is always NULL if zA does
818- ** not end in "*" */
819- assert( pnMatch==0 );
820- }
821- sqlite3_free(toFree);
822- return res;
823+ sqlite3_free(toFree);
824+ return res;
825 }
826
827 /*
828@@ -516,25 +530,26 @@ ** a prefix of B and extra characters on the end of B have minimal additional
829 ** cost.
830 */
831 static void editdistSqlFunc(
832- sqlite3_context *context,
833- int argc,
834- sqlite3_value **argv
835-){
836- int res = editdist1(
837- (const char*)sqlite3_value_text(argv[0]),
838- (const char*)sqlite3_value_text(argv[1]),
839- 0);
840- if( res<0 ){
841- if( res==(-3) ){
842- sqlite3_result_error_nomem(context);
843- }else if( res==(-2) ){
844- sqlite3_result_error(context, "non-ASCII input to editdist()", -1);
845- }else{
846- sqlite3_result_error(context, "NULL input to editdist()", -1);
847+ sqlite3_context *context,
848+ int argc,
849+ sqlite3_value **argv
850+)
851+{
852+ int res = editdist1(
853+ (const char*)sqlite3_value_text(argv[0]),
854+ (const char*)sqlite3_value_text(argv[1]),
855+ 0);
856+ if( res<0 ) {
857+ if( res==(-3) ) {
858+ sqlite3_result_error_nomem(context);
859+ } else if( res==(-2) ) {
860+ sqlite3_result_error(context, "non-ASCII input to editdist()", -1);
861+ } else {
862+ sqlite3_result_error(context, "NULL input to editdist()", -1);
863+ }
864+ } else {
865+ sqlite3_result_int(context, res);
866 }
867- }else{
868- sqlite3_result_int(context, res);
869- }
870 }
871
872 /* End of the fixed-cost edit distance implementation
873@@ -552,30 +567,28 @@ typedef struct EditDist3To EditDist3To;
874 typedef struct EditDist3ToString EditDist3ToString;
875 typedef struct EditDist3Lang EditDist3Lang;
876
877-
878 /*
879 ** An entry in the edit cost table
880 */
881 struct EditDist3Cost {
882- EditDist3Cost *pNext; /* Next cost element */
883- u8 nFrom; /* Number of bytes in aFrom */
884- u8 nTo; /* Number of bytes in aTo */
885- u16 iCost; /* Cost of this transformation */
886- char a[4] ; /* FROM string followed by TO string */
887- /* Additional TO and FROM string bytes appended as necessary */
888+ EditDist3Cost *pNext; /* Next cost element */
889+ u8 nFrom; /* Number of bytes in aFrom */
890+ u8 nTo; /* Number of bytes in aTo */
891+ u16 iCost; /* Cost of this transformation */
892+ char a[4] ; /* FROM string followed by TO string */
893+ /* Additional TO and FROM string bytes appended as necessary */
894 };
895
896 /*
897-** Edit costs for a particular language ID
898+** Edit costs for a particular language ID
899 */
900 struct EditDist3Lang {
901- int iLang; /* Language ID */
902- int iInsCost; /* Default insertion cost */
903- int iDelCost; /* Default deletion cost */
904- int iSubCost; /* Default substitution cost */
905- EditDist3Cost *pCost; /* Costs */
906+ int iLang; /* Language ID */
907+ int iInsCost; /* Default insertion cost */
908+ int iDelCost; /* Default deletion cost */
909+ int iSubCost; /* Default substitution cost */
910+ EditDist3Cost *pCost; /* Costs */
911 };
912-
913
914 /*
915 ** The default EditDist3Lang object, with default costs.
916@@ -586,19 +599,19 @@ /*
917 ** Complete configuration
918 */
919 struct EditDist3Config {
920- int nLang; /* Number of language IDs. Size of a[] */
921- EditDist3Lang *a; /* One for each distinct language ID */
922+ int nLang; /* Number of language IDs. Size of a[] */
923+ EditDist3Lang *a; /* One for each distinct language ID */
924 };
925
926 /*
927 ** Extra information about each character in the FROM string.
928 */
929 struct EditDist3From {
930- int nSubst; /* Number of substitution cost entries */
931- int nDel; /* Number of deletion cost entries */
932- int nByte; /* Number of bytes in this character */
933- EditDist3Cost **apSubst; /* Array of substitution costs for this element */
934- EditDist3Cost **apDel; /* Array of deletion cost entries */
935+ int nSubst; /* Number of substitution cost entries */
936+ int nDel; /* Number of deletion cost entries */
937+ int nByte; /* Number of bytes in this character */
938+ EditDist3Cost **apSubst; /* Array of substitution costs for this element */
939+ EditDist3Cost **apDel; /* Array of deletion cost entries */
940 };
941
942 /*
943@@ -609,66 +622,69 @@ ** In other words, the common case will be to measure the edit distance
944 ** from a single origin string to multiple target strings.
945 */
946 struct EditDist3FromString {
947- char *z; /* The complete text of the FROM string */
948- int n; /* Number of characters in the FROM string */
949- int isPrefix; /* True if ends with '*' character */
950- EditDist3From *a; /* Extra info about each char of the FROM string */
951+ char *z; /* The complete text of the FROM string */
952+ int n; /* Number of characters in the FROM string */
953+ int isPrefix; /* True if ends with '*' character */
954+ EditDist3From *a; /* Extra info about each char of the FROM string */
955 };
956
957 /*
958 ** Extra information about each character in the TO string.
959 */
960 struct EditDist3To {
961- int nIns; /* Number of insertion cost entries */
962- int nByte; /* Number of bytes in this character */
963- EditDist3Cost **apIns; /* Array of deletion cost entries */
964+ int nIns; /* Number of insertion cost entries */
965+ int nByte; /* Number of bytes in this character */
966+ EditDist3Cost **apIns; /* Array of deletion cost entries */
967 };
968
969 /*
970 ** A precompiled FROM string
971 */
972 struct EditDist3ToString {
973- char *z; /* The complete text of the TO string */
974- int n; /* Number of characters in the TO string */
975- EditDist3To *a; /* Extra info about each char of the TO string */
976+ char *z; /* The complete text of the TO string */
977+ int n; /* Number of characters in the TO string */
978+ EditDist3To *a; /* Extra info about each char of the TO string */
979 };
980
981 /*
982 ** Clear or delete an instance of the object that records all edit-distance
983 ** weights.
984 */
985-static void editDist3ConfigClear(EditDist3Config *p){
986- int i;
987- if( p==0 ) return;
988- for(i=0; i<p->nLang; i++){
989- EditDist3Cost *pCost, *pNext;
990- pCost = p->a[i].pCost;
991- while( pCost ){
992- pNext = pCost->pNext;
993- sqlite3_free(pCost);
994- pCost = pNext;
995+static void editDist3ConfigClear(EditDist3Config *p)
996+{
997+ int i;
998+ if( p==0 ) return;
999+ for(i=0; i<p->nLang; i++) {
1000+ EditDist3Cost *pCost, *pNext;
1001+ pCost = p->a[i].pCost;
1002+ while( pCost ) {
1003+ pNext = pCost->pNext;
1004+ sqlite3_free(pCost);
1005+ pCost = pNext;
1006+ }
1007 }
1008- }
1009- sqlite3_free(p->a);
1010- memset(p, 0, sizeof(*p));
1011+ sqlite3_free(p->a);
1012+ memset(p, 0, sizeof(*p));
1013 }
1014-static void editDist3ConfigDelete(void *pIn){
1015- EditDist3Config *p = (EditDist3Config*)pIn;
1016- editDist3ConfigClear(p);
1017- sqlite3_free(p);
1018+static void editDist3ConfigDelete(void *pIn)
1019+{
1020+ EditDist3Config *p = (EditDist3Config*)pIn;
1021+ editDist3ConfigClear(p);
1022+ sqlite3_free(p);
1023 }
1024
1025 /* Compare the FROM values of two EditDist3Cost objects, for sorting.
1026 ** Return negative, zero, or positive if the A is less than, equal to,
1027 ** or greater than B.
1028 */
1029-static int editDist3CostCompare(EditDist3Cost *pA, EditDist3Cost *pB){
1030- int n = pA->nFrom;
1031- int rc;
1032- if( n>pB->nFrom ) n = pB->nFrom;
1033- rc = strncmp(pA->a, pB->a, n);
1034- if( rc==0 ) rc = pA->nFrom - pB->nFrom;
1035- return rc;
1036+static int editDist3CostCompare(EditDist3Cost *pA, EditDist3Cost *pB)
1037+{
1038+ int n = pA->nFrom;
1039+ int rc;
1040+ if( n>pB->nFrom ) n = pB->nFrom;
1041+ rc = strncmp(pA->a, pB->a, n);
1042+ if( rc==0 ) rc = pA->nFrom - pB->nFrom;
1043+ return rc;
1044 }
1045
1046 /*
1047@@ -676,183 +692,195 @@ ** Merge together two sorted lists of EditDist3Cost objects, in order
1048 ** of increasing FROM.
1049 */
1050 static EditDist3Cost *editDist3CostMerge(
1051- EditDist3Cost *pA,
1052- EditDist3Cost *pB
1053-){
1054- EditDist3Cost *pHead = 0;
1055- EditDist3Cost **ppTail = &pHead;
1056- EditDist3Cost *p;
1057- while( pA && pB ){
1058- if( editDist3CostCompare(pA,pB)<=0 ){
1059- p = pA;
1060- pA = pA->pNext;
1061- }else{
1062- p = pB;
1063- pB = pB->pNext;
1064+ EditDist3Cost *pA,
1065+ EditDist3Cost *pB
1066+)
1067+{
1068+ EditDist3Cost *pHead = 0;
1069+ EditDist3Cost **ppTail = &pHead;
1070+ EditDist3Cost *p;
1071+ while( pA && pB ) {
1072+ if( editDist3CostCompare(pA,pB)<=0 ) {
1073+ p = pA;
1074+ pA = pA->pNext;
1075+ } else {
1076+ p = pB;
1077+ pB = pB->pNext;
1078+ }
1079+ *ppTail = p;
1080+ ppTail = &p->pNext;
1081+ }
1082+ if( pA ) {
1083+ *ppTail = pA;
1084+ } else {
1085+ *ppTail = pB;
1086 }
1087- *ppTail = p;
1088- ppTail = &p->pNext;
1089- }
1090- if( pA ){
1091- *ppTail = pA;
1092- }else{
1093- *ppTail = pB;
1094- }
1095- return pHead;
1096+ return pHead;
1097 }
1098
1099 /*
1100 ** Sort a list of EditDist3Cost objects into order of increasing FROM
1101 */
1102-static EditDist3Cost *editDist3CostSort(EditDist3Cost *pList){
1103- EditDist3Cost *ap[60], *p;
1104- int i;
1105- int mx = 0;
1106- ap[0] = 0;
1107- ap[1] = 0;
1108- while( pList ){
1109- p = pList;
1110- pList = p->pNext;
1111- p->pNext = 0;
1112- for(i=0; ap[i]; i++){
1113- p = editDist3CostMerge(ap[i],p);
1114- ap[i] = 0;
1115+static EditDist3Cost *editDist3CostSort(EditDist3Cost *pList)
1116+{
1117+ EditDist3Cost *ap[60], *p;
1118+ int i;
1119+ int mx = 0;
1120+ ap[0] = 0;
1121+ ap[1] = 0;
1122+ while( pList ) {
1123+ p = pList;
1124+ pList = p->pNext;
1125+ p->pNext = 0;
1126+ for(i=0; ap[i]; i++) {
1127+ p = editDist3CostMerge(ap[i],p);
1128+ ap[i] = 0;
1129+ }
1130+ ap[i] = p;
1131+ if( i>mx ) {
1132+ mx = i;
1133+ ap[i+1] = 0;
1134+ }
1135 }
1136- ap[i] = p;
1137- if( i>mx ){
1138- mx = i;
1139- ap[i+1] = 0;
1140+ p = 0;
1141+ for(i=0; i<=mx; i++) {
1142+ if( ap[i] ) p = editDist3CostMerge(p,ap[i]);
1143 }
1144- }
1145- p = 0;
1146- for(i=0; i<=mx; i++){
1147- if( ap[i] ) p = editDist3CostMerge(p,ap[i]);
1148- }
1149- return p;
1150+ return p;
1151 }
1152
1153 /*
1154 ** Load all edit-distance weights from a table.
1155 */
1156 static int editDist3ConfigLoad(
1157- EditDist3Config *p, /* The edit distance configuration to load */
1158- sqlite3 *db, /* Load from this database */
1159- const char *zTable /* Name of the table from which to load */
1160-){
1161- sqlite3_stmt *pStmt;
1162- int rc, rc2;
1163- char *zSql;
1164- int iLangPrev = -9999;
1165- EditDist3Lang *pLang = 0;
1166+ EditDist3Config *p, /* The edit distance configuration to load */
1167+ sqlite3 *db, /* Load from this database */
1168+ const char *zTable /* Name of the table from which to load */
1169+)
1170+{
1171+ sqlite3_stmt *pStmt;
1172+ int rc, rc2;
1173+ char *zSql;
1174+ int iLangPrev = -9999;
1175+ EditDist3Lang *pLang = 0;
1176
1177- zSql = sqlite3_mprintf("SELECT iLang, cFrom, cTo, iCost"
1178- " FROM \"%w\" WHERE iLang>=0 ORDER BY iLang", zTable);
1179- if( zSql==0 ) return SQLITE_NOMEM;
1180- rc = sqlite3_prepare(db, zSql, -1, &pStmt, 0);
1181- sqlite3_free(zSql);
1182- if( rc ) return rc;
1183- editDist3ConfigClear(p);
1184- while( sqlite3_step(pStmt)==SQLITE_ROW ){
1185- int iLang = sqlite3_column_int(pStmt, 0);
1186- const char *zFrom = (const char*)sqlite3_column_text(pStmt, 1);
1187- int nFrom = zFrom ? sqlite3_column_bytes(pStmt, 1) : 0;
1188- const char *zTo = (const char*)sqlite3_column_text(pStmt, 2);
1189- int nTo = zTo ? sqlite3_column_bytes(pStmt, 2) : 0;
1190- int iCost = sqlite3_column_int(pStmt, 3);
1191+ zSql = sqlite3_mprintf("SELECT iLang, cFrom, cTo, iCost"
1192+ " FROM \"%w\" WHERE iLang>=0 ORDER BY iLang", zTable);
1193+ if( zSql==0 ) return SQLITE_NOMEM;
1194+ rc = sqlite3_prepare(db, zSql, -1, &pStmt, 0);
1195+ sqlite3_free(zSql);
1196+ if( rc ) return rc;
1197+ editDist3ConfigClear(p);
1198+ while( sqlite3_step(pStmt)==SQLITE_ROW ) {
1199+ int iLang = sqlite3_column_int(pStmt, 0);
1200+ const char *zFrom = (const char*)sqlite3_column_text(pStmt, 1);
1201+ int nFrom = zFrom ? sqlite3_column_bytes(pStmt, 1) : 0;
1202+ const char *zTo = (const char*)sqlite3_column_text(pStmt, 2);
1203+ int nTo = zTo ? sqlite3_column_bytes(pStmt, 2) : 0;
1204+ int iCost = sqlite3_column_int(pStmt, 3);
1205
1206- assert( zFrom!=0 || nFrom==0 );
1207- assert( zTo!=0 || nTo==0 );
1208- if( nFrom>100 || nTo>100 ) continue;
1209- if( iCost<0 ) continue;
1210- if( iCost>=10000 ) continue; /* Costs above 10K are considered infinite */
1211- if( pLang==0 || iLang!=iLangPrev ){
1212- EditDist3Lang *pNew;
1213- pNew = sqlite3_realloc64(p->a, (p->nLang+1)*sizeof(p->a[0]));
1214- if( pNew==0 ){ rc = SQLITE_NOMEM; break; }
1215- p->a = pNew;
1216- pLang = &p->a[p->nLang];
1217- p->nLang++;
1218- pLang->iLang = iLang;
1219- pLang->iInsCost = 100;
1220- pLang->iDelCost = 100;
1221- pLang->iSubCost = 150;
1222- pLang->pCost = 0;
1223- iLangPrev = iLang;
1224+ assert( zFrom!=0 || nFrom==0 );
1225+ assert( zTo!=0 || nTo==0 );
1226+ if( nFrom>100 || nTo>100 ) continue;
1227+ if( iCost<0 ) continue;
1228+ if( iCost>=10000 ) continue; /* Costs above 10K are considered infinite */
1229+ if( pLang==0 || iLang!=iLangPrev ) {
1230+ EditDist3Lang *pNew;
1231+ pNew = sqlite3_realloc64(p->a, (p->nLang+1)*sizeof(p->a[0]));
1232+ if( pNew==0 ) {
1233+ rc = SQLITE_NOMEM;
1234+ break;
1235+ }
1236+ p->a = pNew;
1237+ pLang = &p->a[p->nLang];
1238+ p->nLang++;
1239+ pLang->iLang = iLang;
1240+ pLang->iInsCost = 100;
1241+ pLang->iDelCost = 100;
1242+ pLang->iSubCost = 150;
1243+ pLang->pCost = 0;
1244+ iLangPrev = iLang;
1245+ }
1246+ if( nFrom==1 && zFrom[0]=='?' && nTo==0 ) {
1247+ pLang->iDelCost = iCost;
1248+ } else if( nFrom==0 && nTo==1 && zTo[0]=='?' ) {
1249+ pLang->iInsCost = iCost;
1250+ } else if( nFrom==1 && nTo==1 && zFrom[0]=='?' && zTo[0]=='?' ) {
1251+ pLang->iSubCost = iCost;
1252+ } else {
1253+ EditDist3Cost *pCost;
1254+ int nExtra = nFrom + nTo - 4;
1255+ if( nExtra<0 ) nExtra = 0;
1256+ pCost = sqlite3_malloc64( sizeof(*pCost) + nExtra );
1257+ if( pCost==0 ) {
1258+ rc = SQLITE_NOMEM;
1259+ break;
1260+ }
1261+ pCost->nFrom = (u8)nFrom;
1262+ pCost->nTo = (u8)nTo;
1263+ pCost->iCost = (u16)iCost;
1264+ memcpy(pCost->a, zFrom, nFrom);
1265+ memcpy(pCost->a + nFrom, zTo, nTo);
1266+ pCost->pNext = pLang->pCost;
1267+ pLang->pCost = pCost;
1268+ }
1269 }
1270- if( nFrom==1 && zFrom[0]=='?' && nTo==0 ){
1271- pLang->iDelCost = iCost;
1272- }else if( nFrom==0 && nTo==1 && zTo[0]=='?' ){
1273- pLang->iInsCost = iCost;
1274- }else if( nFrom==1 && nTo==1 && zFrom[0]=='?' && zTo[0]=='?' ){
1275- pLang->iSubCost = iCost;
1276- }else{
1277- EditDist3Cost *pCost;
1278- int nExtra = nFrom + nTo - 4;
1279- if( nExtra<0 ) nExtra = 0;
1280- pCost = sqlite3_malloc64( sizeof(*pCost) + nExtra );
1281- if( pCost==0 ){ rc = SQLITE_NOMEM; break; }
1282- pCost->nFrom = (u8)nFrom;
1283- pCost->nTo = (u8)nTo;
1284- pCost->iCost = (u16)iCost;
1285- memcpy(pCost->a, zFrom, nFrom);
1286- memcpy(pCost->a + nFrom, zTo, nTo);
1287- pCost->pNext = pLang->pCost;
1288- pLang->pCost = pCost;
1289- }
1290- }
1291- rc2 = sqlite3_finalize(pStmt);
1292- if( rc==SQLITE_OK ) rc = rc2;
1293- if( rc==SQLITE_OK ){
1294- int iLang;
1295- for(iLang=0; iLang<p->nLang; iLang++){
1296- p->a[iLang].pCost = editDist3CostSort(p->a[iLang].pCost);
1297+ rc2 = sqlite3_finalize(pStmt);
1298+ if( rc==SQLITE_OK ) rc = rc2;
1299+ if( rc==SQLITE_OK ) {
1300+ int iLang;
1301+ for(iLang=0; iLang<p->nLang; iLang++) {
1302+ p->a[iLang].pCost = editDist3CostSort(p->a[iLang].pCost);
1303+ }
1304 }
1305- }
1306- return rc;
1307+ return rc;
1308 }
1309
1310 /*
1311 ** Return the length (in bytes) of a utf-8 character. Or return a maximum
1312 ** of N.
1313 */
1314-static int utf8Len(unsigned char c, int N){
1315- int len = 1;
1316- if( c>0x7f ){
1317- if( (c&0xe0)==0xc0 ){
1318- len = 2;
1319- }else if( (c&0xf0)==0xe0 ){
1320- len = 3;
1321- }else{
1322- len = 4;
1323+static int utf8Len(unsigned char c, int N)
1324+{
1325+ int len = 1;
1326+ if( c>0x7f ) {
1327+ if( (c&0xe0)==0xc0 ) {
1328+ len = 2;
1329+ } else if( (c&0xf0)==0xe0 ) {
1330+ len = 3;
1331+ } else {
1332+ len = 4;
1333+ }
1334 }
1335- }
1336- if( len>N ) len = N;
1337- return len;
1338+ if( len>N ) len = N;
1339+ return len;
1340 }
1341
1342 /*
1343 ** Return TRUE (non-zero) if the To side of the given cost matches
1344 ** the given string.
1345 */
1346-static int matchTo(EditDist3Cost *p, const char *z, int n){
1347- assert( n>0 );
1348- if( p->a[p->nFrom]!=z[0] ) return 0;
1349- if( p->nTo>n ) return 0;
1350- if( strncmp(p->a+p->nFrom, z, p->nTo)!=0 ) return 0;
1351- return 1;
1352+static int matchTo(EditDist3Cost *p, const char *z, int n)
1353+{
1354+ assert( n>0 );
1355+ if( p->a[p->nFrom]!=z[0] ) return 0;
1356+ if( p->nTo>n ) return 0;
1357+ if( strncmp(p->a+p->nFrom, z, p->nTo)!=0 ) return 0;
1358+ return 1;
1359 }
1360
1361 /*
1362 ** Return TRUE (non-zero) if the From side of the given cost matches
1363 ** the given string.
1364 */
1365-static int matchFrom(EditDist3Cost *p, const char *z, int n){
1366- assert( p->nFrom<=n );
1367- if( p->nFrom ){
1368- if( p->a[0]!=z[0] ) return 0;
1369- if( strncmp(p->a, z, p->nFrom)!=0 ) return 0;
1370- }
1371- return 1;
1372+static int matchFrom(EditDist3Cost *p, const char *z, int n)
1373+{
1374+ assert( p->nFrom<=n );
1375+ if( p->nFrom ) {
1376+ if( p->a[0]!=z[0] ) return 0;
1377+ if( strncmp(p->a, z, p->nFrom)!=0 ) return 0;
1378+ }
1379+ return 1;
1380 }
1381
1382 /*
1383@@ -860,92 +888,95 @@ ** Return TRUE (non-zero) of the next FROM character and the next TO
1384 ** character are the same.
1385 */
1386 static int matchFromTo(
1387- EditDist3FromString *pStr, /* Left hand string */
1388- int n1, /* Index of comparison character on the left */
1389- const char *z2, /* Right-handl comparison character */
1390- int n2 /* Bytes remaining in z2[] */
1391-){
1392- int b1 = pStr->a[n1].nByte;
1393- if( b1>n2 ) return 0;
1394- assert( b1>0 );
1395- if( pStr->z[n1]!=z2[0] ) return 0;
1396- if( strncmp(pStr->z+n1, z2, b1)!=0 ) return 0;
1397- return 1;
1398+ EditDist3FromString *pStr, /* Left hand string */
1399+ int n1, /* Index of comparison character on the left */
1400+ const char *z2, /* Right-handl comparison character */
1401+ int n2 /* Bytes remaining in z2[] */
1402+)
1403+{
1404+ int b1 = pStr->a[n1].nByte;
1405+ if( b1>n2 ) return 0;
1406+ assert( b1>0 );
1407+ if( pStr->z[n1]!=z2[0] ) return 0;
1408+ if( strncmp(pStr->z+n1, z2, b1)!=0 ) return 0;
1409+ return 1;
1410 }
1411
1412 /*
1413 ** Delete an EditDist3FromString objecct
1414 */
1415-static void editDist3FromStringDelete(EditDist3FromString *p){
1416- int i;
1417- if( p ){
1418- for(i=0; i<p->n; i++){
1419- sqlite3_free(p->a[i].apDel);
1420- sqlite3_free(p->a[i].apSubst);
1421+static void editDist3FromStringDelete(EditDist3FromString *p)
1422+{
1423+ int i;
1424+ if( p ) {
1425+ for(i=0; i<p->n; i++) {
1426+ sqlite3_free(p->a[i].apDel);
1427+ sqlite3_free(p->a[i].apSubst);
1428+ }
1429+ sqlite3_free(p);
1430 }
1431- sqlite3_free(p);
1432- }
1433 }
1434
1435 /*
1436 ** Create a EditDist3FromString object.
1437 */
1438 static EditDist3FromString *editDist3FromStringNew(
1439- const EditDist3Lang *pLang,
1440- const char *z,
1441- int n
1442-){
1443- EditDist3FromString *pStr;
1444- EditDist3Cost *p;
1445- int i;
1446-
1447- if( z==0 ) return 0;
1448- if( n<0 ) n = (int)strlen(z);
1449- pStr = sqlite3_malloc64( sizeof(*pStr) + sizeof(pStr->a[0])*n + n + 1 );
1450- if( pStr==0 ) return 0;
1451- pStr->a = (EditDist3From*)&pStr[1];
1452- memset(pStr->a, 0, sizeof(pStr->a[0])*n);
1453- pStr->n = n;
1454- pStr->z = (char*)&pStr->a[n];
1455- memcpy(pStr->z, z, n+1);
1456- if( n && z[n-1]=='*' ){
1457- pStr->isPrefix = 1;
1458- n--;
1459- pStr->n--;
1460- pStr->z[n] = 0;
1461- }else{
1462- pStr->isPrefix = 0;
1463- }
1464+ const EditDist3Lang *pLang,
1465+ const char *z,
1466+ int n
1467+)
1468+{
1469+ EditDist3FromString *pStr;
1470+ EditDist3Cost *p;
1471+ int i;
1472
1473- for(i=0; i<n; i++){
1474- EditDist3From *pFrom = &pStr->a[i];
1475- memset(pFrom, 0, sizeof(*pFrom));
1476- pFrom->nByte = utf8Len((unsigned char)z[i], n-i);
1477- for(p=pLang->pCost; p; p=p->pNext){
1478- EditDist3Cost **apNew;
1479- if( i+p->nFrom>n ) continue;
1480- if( matchFrom(p, z+i, n-i)==0 ) continue;
1481- if( p->nTo==0 ){
1482- apNew = sqlite3_realloc64(pFrom->apDel,
1483- sizeof(*apNew)*(pFrom->nDel+1));
1484- if( apNew==0 ) break;
1485- pFrom->apDel = apNew;
1486- apNew[pFrom->nDel++] = p;
1487- }else{
1488- apNew = sqlite3_realloc64(pFrom->apSubst,
1489- sizeof(*apNew)*(pFrom->nSubst+1));
1490- if( apNew==0 ) break;
1491- pFrom->apSubst = apNew;
1492- apNew[pFrom->nSubst++] = p;
1493- }
1494+ if( z==0 ) return 0;
1495+ if( n<0 ) n = (int)strlen(z);
1496+ pStr = sqlite3_malloc64( sizeof(*pStr) + sizeof(pStr->a[0])*n + n + 1 );
1497+ if( pStr==0 ) return 0;
1498+ pStr->a = (EditDist3From*)&pStr[1];
1499+ memset(pStr->a, 0, sizeof(pStr->a[0])*n);
1500+ pStr->n = n;
1501+ pStr->z = (char*)&pStr->a[n];
1502+ memcpy(pStr->z, z, n+1);
1503+ if( n && z[n-1]=='*' ) {
1504+ pStr->isPrefix = 1;
1505+ n--;
1506+ pStr->n--;
1507+ pStr->z[n] = 0;
1508+ } else {
1509+ pStr->isPrefix = 0;
1510 }
1511- if( p ){
1512- editDist3FromStringDelete(pStr);
1513- pStr = 0;
1514- break;
1515+
1516+ for(i=0; i<n; i++) {
1517+ EditDist3From *pFrom = &pStr->a[i];
1518+ memset(pFrom, 0, sizeof(*pFrom));
1519+ pFrom->nByte = utf8Len((unsigned char)z[i], n-i);
1520+ for(p=pLang->pCost; p; p=p->pNext) {
1521+ EditDist3Cost **apNew;
1522+ if( i+p->nFrom>n ) continue;
1523+ if( matchFrom(p, z+i, n-i)==0 ) continue;
1524+ if( p->nTo==0 ) {
1525+ apNew = sqlite3_realloc64(pFrom->apDel,
1526+ sizeof(*apNew)*(pFrom->nDel+1));
1527+ if( apNew==0 ) break;
1528+ pFrom->apDel = apNew;
1529+ apNew[pFrom->nDel++] = p;
1530+ } else {
1531+ apNew = sqlite3_realloc64(pFrom->apSubst,
1532+ sizeof(*apNew)*(pFrom->nSubst+1));
1533+ if( apNew==0 ) break;
1534+ pFrom->apSubst = apNew;
1535+ apNew[pFrom->nSubst++] = p;
1536+ }
1537+ }
1538+ if( p ) {
1539+ editDist3FromStringDelete(pStr);
1540+ pStr = 0;
1541+ break;
1542+ }
1543 }
1544- }
1545- return pStr;
1546+ return pStr;
1547 }
1548
1549 /*
1550@@ -953,20 +984,21 @@ ** Update entry m[i] such that it is the minimum of its current value
1551 ** and m[j]+iCost.
1552 */
1553 static void updateCost(
1554- unsigned int *m,
1555- int i,
1556- int j,
1557- int iCost
1558-){
1559- unsigned int b;
1560- assert( iCost>=0 );
1561- assert( iCost<10000 );
1562- b = m[j] + iCost;
1563- if( b<m[i] ) m[i] = b;
1564+ unsigned int *m,
1565+ int i,
1566+ int j,
1567+ int iCost
1568+)
1569+{
1570+ unsigned int b;
1571+ assert( iCost>=0 );
1572+ assert( iCost<10000 );
1573+ b = m[j] + iCost;
1574+ if( b<m[i] ) m[i] = b;
1575 }
1576
1577 /*
1578-** How much stack space (int bytes) to use for Wagner matrix in
1579+** How much stack space (int bytes) to use for Wagner matrix in
1580 ** editDist3Core(). If more space than this is required, the entire
1581 ** matrix is taken from the heap. To reduce the load on the memory
1582 ** allocator, make this value as large as practical for the
1583@@ -984,178 +1016,180 @@ ** If pnMatch is not NULL, then *pnMatch is set to the number of characters
1584 ** (not bytes) in z2 that matched the search pattern in *pFrom. If pFrom does
1585 ** not contain the pattern for a prefix-search, then this is always the number
1586 ** of characters in z2. If pFrom does contain a prefix search pattern, then
1587-** it is the number of characters in the prefix of z2 that was deemed to
1588+** it is the number of characters in the prefix of z2 that was deemed to
1589 ** match pFrom.
1590 */
1591 static int editDist3Core(
1592- EditDist3FromString *pFrom, /* The FROM string */
1593- const char *z2, /* The TO string */
1594- int n2, /* Length of the TO string */
1595- const EditDist3Lang *pLang, /* Edit weights for a particular language ID */
1596- int *pnMatch /* OUT: Characters in matched prefix */
1597-){
1598- int k, n;
1599- int i1, b1;
1600- int i2, b2;
1601- EditDist3FromString f = *pFrom;
1602- EditDist3To *a2;
1603- unsigned int *m;
1604- unsigned int *pToFree;
1605- int szRow;
1606- EditDist3Cost *p;
1607- int res;
1608- sqlite3_uint64 nByte;
1609- unsigned int stackSpace[SQLITE_SPELLFIX_STACKALLOC_SZ/sizeof(unsigned int)];
1610+ EditDist3FromString *pFrom, /* The FROM string */
1611+ const char *z2, /* The TO string */
1612+ int n2, /* Length of the TO string */
1613+ const EditDist3Lang *pLang, /* Edit weights for a particular language ID */
1614+ int *pnMatch /* OUT: Characters in matched prefix */
1615+)
1616+{
1617+ int k, n;
1618+ int i1, b1;
1619+ int i2, b2;
1620+ EditDist3FromString f = *pFrom;
1621+ EditDist3To *a2;
1622+ unsigned int *m;
1623+ unsigned int *pToFree;
1624+ int szRow;
1625+ EditDist3Cost *p;
1626+ int res;
1627+ sqlite3_uint64 nByte;
1628+ unsigned int stackSpace[SQLITE_SPELLFIX_STACKALLOC_SZ/sizeof(unsigned int)];
1629
1630- /* allocate the Wagner matrix and the aTo[] array for the TO string */
1631- n = (f.n+1)*(n2+1);
1632- n = (n+1)&~1;
1633- nByte = n*sizeof(m[0]) + sizeof(a2[0])*n2;
1634- if( nByte<=sizeof(stackSpace) ){
1635- m = stackSpace;
1636- pToFree = 0;
1637- }else{
1638- m = pToFree = sqlite3_malloc64( nByte );
1639- if( m==0 ) return -1; /* Out of memory */
1640- }
1641- a2 = (EditDist3To*)&m[n];
1642- memset(a2, 0, sizeof(a2[0])*n2);
1643+ /* allocate the Wagner matrix and the aTo[] array for the TO string */
1644+ n = (f.n+1)*(n2+1);
1645+ n = (n+1)&~1;
1646+ nByte = n*sizeof(m[0]) + sizeof(a2[0])*n2;
1647+ if( nByte<=sizeof(stackSpace) ) {
1648+ m = stackSpace;
1649+ pToFree = 0;
1650+ } else {
1651+ m = pToFree = sqlite3_malloc64( nByte );
1652+ if( m==0 ) return -1; /* Out of memory */
1653+ }
1654+ a2 = (EditDist3To*)&m[n];
1655+ memset(a2, 0, sizeof(a2[0])*n2);
1656
1657- /* Fill in the a1[] matrix for all characters of the TO string */
1658- for(i2=0; i2<n2; i2++){
1659- a2[i2].nByte = utf8Len((unsigned char)z2[i2], n2-i2);
1660- for(p=pLang->pCost; p; p=p->pNext){
1661- EditDist3Cost **apNew;
1662- if( p->nFrom>0 ) break;
1663- if( i2+p->nTo>n2 ) continue;
1664- if( p->a[0]>z2[i2] ) break;
1665- if( matchTo(p, z2+i2, n2-i2)==0 ) continue;
1666- a2[i2].nIns++;
1667- apNew = sqlite3_realloc64(a2[i2].apIns, sizeof(*apNew)*a2[i2].nIns);
1668- if( apNew==0 ){
1669- res = -1; /* Out of memory */
1670- goto editDist3Abort;
1671- }
1672- a2[i2].apIns = apNew;
1673- a2[i2].apIns[a2[i2].nIns-1] = p;
1674+ /* Fill in the a1[] matrix for all characters of the TO string */
1675+ for(i2=0; i2<n2; i2++) {
1676+ a2[i2].nByte = utf8Len((unsigned char)z2[i2], n2-i2);
1677+ for(p=pLang->pCost; p; p=p->pNext) {
1678+ EditDist3Cost **apNew;
1679+ if( p->nFrom>0 ) break;
1680+ if( i2+p->nTo>n2 ) continue;
1681+ if( p->a[0]>z2[i2] ) break;
1682+ if( matchTo(p, z2+i2, n2-i2)==0 ) continue;
1683+ a2[i2].nIns++;
1684+ apNew = sqlite3_realloc64(a2[i2].apIns, sizeof(*apNew)*a2[i2].nIns);
1685+ if( apNew==0 ) {
1686+ res = -1; /* Out of memory */
1687+ goto editDist3Abort;
1688+ }
1689+ a2[i2].apIns = apNew;
1690+ a2[i2].apIns[a2[i2].nIns-1] = p;
1691+ }
1692 }
1693- }
1694
1695- /* Prepare to compute the minimum edit distance */
1696- szRow = f.n+1;
1697- memset(m, 0x01, (n2+1)*szRow*sizeof(m[0]));
1698- m[0] = 0;
1699+ /* Prepare to compute the minimum edit distance */
1700+ szRow = f.n+1;
1701+ memset(m, 0x01, (n2+1)*szRow*sizeof(m[0]));
1702+ m[0] = 0;
1703
1704- /* First fill in the top-row of the matrix with FROM deletion costs */
1705- for(i1=0; i1<f.n; i1 += b1){
1706- b1 = f.a[i1].nByte;
1707- updateCost(m, i1+b1, i1, pLang->iDelCost);
1708- for(k=0; k<f.a[i1].nDel; k++){
1709- p = f.a[i1].apDel[k];
1710- updateCost(m, i1+p->nFrom, i1, p->iCost);
1711+ /* First fill in the top-row of the matrix with FROM deletion costs */
1712+ for(i1=0; i1<f.n; i1 += b1) {
1713+ b1 = f.a[i1].nByte;
1714+ updateCost(m, i1+b1, i1, pLang->iDelCost);
1715+ for(k=0; k<f.a[i1].nDel; k++) {
1716+ p = f.a[i1].apDel[k];
1717+ updateCost(m, i1+p->nFrom, i1, p->iCost);
1718+ }
1719 }
1720- }
1721
1722- /* Fill in all subsequent rows, top-to-bottom, left-to-right */
1723- for(i2=0; i2<n2; i2 += b2){
1724- int rx; /* Starting index for current row */
1725- int rxp; /* Starting index for previous row */
1726- b2 = a2[i2].nByte;
1727- rx = szRow*(i2+b2);
1728- rxp = szRow*i2;
1729- updateCost(m, rx, rxp, pLang->iInsCost);
1730- for(k=0; k<a2[i2].nIns; k++){
1731- p = a2[i2].apIns[k];
1732- updateCost(m, szRow*(i2+p->nTo), rxp, p->iCost);
1733- }
1734- for(i1=0; i1<f.n; i1+=b1){
1735- int cx; /* Index of current cell */
1736- int cxp; /* Index of cell immediately to the left */
1737- int cxd; /* Index of cell to the left and one row above */
1738- int cxu; /* Index of cell immediately above */
1739- b1 = f.a[i1].nByte;
1740- cxp = rx + i1;
1741- cx = cxp + b1;
1742- cxd = rxp + i1;
1743- cxu = cxd + b1;
1744- updateCost(m, cx, cxp, pLang->iDelCost);
1745- for(k=0; k<f.a[i1].nDel; k++){
1746- p = f.a[i1].apDel[k];
1747- updateCost(m, cxp+p->nFrom, cxp, p->iCost);
1748- }
1749- updateCost(m, cx, cxu, pLang->iInsCost);
1750- if( matchFromTo(&f, i1, z2+i2, n2-i2) ){
1751- updateCost(m, cx, cxd, 0);
1752- }
1753- updateCost(m, cx, cxd, pLang->iSubCost);
1754- for(k=0; k<f.a[i1].nSubst; k++){
1755- p = f.a[i1].apSubst[k];
1756- if( matchTo(p, z2+i2, n2-i2) ){
1757- updateCost(m, cxd+p->nFrom+szRow*p->nTo, cxd, p->iCost);
1758+ /* Fill in all subsequent rows, top-to-bottom, left-to-right */
1759+ for(i2=0; i2<n2; i2 += b2) {
1760+ int rx; /* Starting index for current row */
1761+ int rxp; /* Starting index for previous row */
1762+ b2 = a2[i2].nByte;
1763+ rx = szRow*(i2+b2);
1764+ rxp = szRow*i2;
1765+ updateCost(m, rx, rxp, pLang->iInsCost);
1766+ for(k=0; k<a2[i2].nIns; k++) {
1767+ p = a2[i2].apIns[k];
1768+ updateCost(m, szRow*(i2+p->nTo), rxp, p->iCost);
1769+ }
1770+ for(i1=0; i1<f.n; i1+=b1) {
1771+ int cx; /* Index of current cell */
1772+ int cxp; /* Index of cell immediately to the left */
1773+ int cxd; /* Index of cell to the left and one row above */
1774+ int cxu; /* Index of cell immediately above */
1775+ b1 = f.a[i1].nByte;
1776+ cxp = rx + i1;
1777+ cx = cxp + b1;
1778+ cxd = rxp + i1;
1779+ cxu = cxd + b1;
1780+ updateCost(m, cx, cxp, pLang->iDelCost);
1781+ for(k=0; k<f.a[i1].nDel; k++) {
1782+ p = f.a[i1].apDel[k];
1783+ updateCost(m, cxp+p->nFrom, cxp, p->iCost);
1784+ }
1785+ updateCost(m, cx, cxu, pLang->iInsCost);
1786+ if( matchFromTo(&f, i1, z2+i2, n2-i2) ) {
1787+ updateCost(m, cx, cxd, 0);
1788+ }
1789+ updateCost(m, cx, cxd, pLang->iSubCost);
1790+ for(k=0; k<f.a[i1].nSubst; k++) {
1791+ p = f.a[i1].apSubst[k];
1792+ if( matchTo(p, z2+i2, n2-i2) ) {
1793+ updateCost(m, cxd+p->nFrom+szRow*p->nTo, cxd, p->iCost);
1794+ }
1795+ }
1796 }
1797- }
1798 }
1799- }
1800
1801 #if 0 /* Enable for debugging */
1802- printf(" ^");
1803- for(i1=0; i1<f.n; i1++) printf(" %c-%2x", f.z[i1], f.z[i1]&0xff);
1804- printf("\n ^:");
1805- for(i1=0; i1<szRow; i1++){
1806- int v = m[i1];
1807- if( v>9999 ) printf(" ****");
1808- else printf(" %4d", v);
1809- }
1810- printf("\n");
1811- for(i2=0; i2<n2; i2++){
1812- printf("%c-%02x:", z2[i2], z2[i2]&0xff);
1813- for(i1=0; i1<szRow; i1++){
1814- int v = m[(i2+1)*szRow+i1];
1815- if( v>9999 ) printf(" ****");
1816- else printf(" %4d", v);
1817+ printf(" ^");
1818+ for(i1=0; i1<f.n; i1++) printf(" %c-%2x", f.z[i1], f.z[i1]&0xff);
1819+ printf("\n ^:");
1820+ for(i1=0; i1<szRow; i1++) {
1821+ int v = m[i1];
1822+ if( v>9999 ) printf(" ****");
1823+ else printf(" %4d", v);
1824 }
1825 printf("\n");
1826- }
1827+ for(i2=0; i2<n2; i2++) {
1828+ printf("%c-%02x:", z2[i2], z2[i2]&0xff);
1829+ for(i1=0; i1<szRow; i1++) {
1830+ int v = m[(i2+1)*szRow+i1];
1831+ if( v>9999 ) printf(" ****");
1832+ else printf(" %4d", v);
1833+ }
1834+ printf("\n");
1835+ }
1836 #endif
1837
1838- /* Free memory allocations and return the result */
1839- res = (int)m[szRow*(n2+1)-1];
1840- n = n2;
1841- if( f.isPrefix ){
1842- for(i2=1; i2<=n2; i2++){
1843- int b = m[szRow*i2-1];
1844- if( b<=res ){
1845- res = b;
1846- n = i2 - 1;
1847- }
1848+ /* Free memory allocations and return the result */
1849+ res = (int)m[szRow*(n2+1)-1];
1850+ n = n2;
1851+ if( f.isPrefix ) {
1852+ for(i2=1; i2<=n2; i2++) {
1853+ int b = m[szRow*i2-1];
1854+ if( b<=res ) {
1855+ res = b;
1856+ n = i2 - 1;
1857+ }
1858+ }
1859 }
1860- }
1861- if( pnMatch ){
1862- int nExtra = 0;
1863- for(k=0; k<n; k++){
1864- if( (z2[k] & 0xc0)==0x80 ) nExtra++;
1865+ if( pnMatch ) {
1866+ int nExtra = 0;
1867+ for(k=0; k<n; k++) {
1868+ if( (z2[k] & 0xc0)==0x80 ) nExtra++;
1869+ }
1870+ *pnMatch = n - nExtra;
1871 }
1872- *pnMatch = n - nExtra;
1873- }
1874
1875 editDist3Abort:
1876- for(i2=0; i2<n2; i2++) sqlite3_free(a2[i2].apIns);
1877- sqlite3_free(pToFree);
1878- return res;
1879+ for(i2=0; i2<n2; i2++) sqlite3_free(a2[i2].apIns);
1880+ sqlite3_free(pToFree);
1881+ return res;
1882 }
1883
1884 /*
1885 ** Get an appropriate EditDist3Lang object.
1886 */
1887 static const EditDist3Lang *editDist3FindLang(
1888- EditDist3Config *pConfig,
1889- int iLang
1890-){
1891- int i;
1892- for(i=0; i<pConfig->nLang; i++){
1893- if( pConfig->a[i].iLang==iLang ) return &pConfig->a[i];
1894- }
1895- return &editDist3Lang;
1896+ EditDist3Config *pConfig,
1897+ int iLang
1898+)
1899+{
1900+ int i;
1901+ for(i=0; i<pConfig->nLang; i++) {
1902+ if( pConfig->a[i].iLang==iLang ) return &pConfig->a[i];
1903+ }
1904+ return &editDist3Lang;
1905 }
1906
1907 /*
1908@@ -1168,66 +1202,68 @@ **
1909 ** The second form loads edit weights into memory from a table.
1910 */
1911 static void editDist3SqlFunc(
1912- sqlite3_context *context,
1913- int argc,
1914- sqlite3_value **argv
1915-){
1916- EditDist3Config *pConfig = (EditDist3Config*)sqlite3_user_data(context);
1917- sqlite3 *db = sqlite3_context_db_handle(context);
1918- int rc;
1919- if( argc==1 ){
1920- const char *zTable = (const char*)sqlite3_value_text(argv[0]);
1921- rc = editDist3ConfigLoad(pConfig, db, zTable);
1922- if( rc ) sqlite3_result_error_code(context, rc);
1923- }else{
1924- const char *zA = (const char*)sqlite3_value_text(argv[0]);
1925- const char *zB = (const char*)sqlite3_value_text(argv[1]);
1926- int nA = sqlite3_value_bytes(argv[0]);
1927- int nB = sqlite3_value_bytes(argv[1]);
1928- int iLang = argc==3 ? sqlite3_value_int(argv[2]) : 0;
1929- const EditDist3Lang *pLang = editDist3FindLang(pConfig, iLang);
1930- EditDist3FromString *pFrom;
1931- int dist;
1932+ sqlite3_context *context,
1933+ int argc,
1934+ sqlite3_value **argv
1935+)
1936+{
1937+ EditDist3Config *pConfig = (EditDist3Config*)sqlite3_user_data(context);
1938+ sqlite3 *db = sqlite3_context_db_handle(context);
1939+ int rc;
1940+ if( argc==1 ) {
1941+ const char *zTable = (const char*)sqlite3_value_text(argv[0]);
1942+ rc = editDist3ConfigLoad(pConfig, db, zTable);
1943+ if( rc ) sqlite3_result_error_code(context, rc);
1944+ } else {
1945+ const char *zA = (const char*)sqlite3_value_text(argv[0]);
1946+ const char *zB = (const char*)sqlite3_value_text(argv[1]);
1947+ int nA = sqlite3_value_bytes(argv[0]);
1948+ int nB = sqlite3_value_bytes(argv[1]);
1949+ int iLang = argc==3 ? sqlite3_value_int(argv[2]) : 0;
1950+ const EditDist3Lang *pLang = editDist3FindLang(pConfig, iLang);
1951+ EditDist3FromString *pFrom;
1952+ int dist;
1953
1954- pFrom = editDist3FromStringNew(pLang, zA, nA);
1955- if( pFrom==0 ){
1956- sqlite3_result_error_nomem(context);
1957- return;
1958- }
1959- dist = editDist3Core(pFrom, zB, nB, pLang, 0);
1960- editDist3FromStringDelete(pFrom);
1961- if( dist==(-1) ){
1962- sqlite3_result_error_nomem(context);
1963- }else{
1964- sqlite3_result_int(context, dist);
1965+ pFrom = editDist3FromStringNew(pLang, zA, nA);
1966+ if( pFrom==0 ) {
1967+ sqlite3_result_error_nomem(context);
1968+ return;
1969+ }
1970+ dist = editDist3Core(pFrom, zB, nB, pLang, 0);
1971+ editDist3FromStringDelete(pFrom);
1972+ if( dist==(-1) ) {
1973+ sqlite3_result_error_nomem(context);
1974+ } else {
1975+ sqlite3_result_int(context, dist);
1976+ }
1977 }
1978- }
1979 }
1980
1981 /*
1982 ** Register the editDist3 function with SQLite
1983 */
1984-static int editDist3Install(sqlite3 *db){
1985- int rc;
1986- EditDist3Config *pConfig = sqlite3_malloc64( sizeof(*pConfig) );
1987- if( pConfig==0 ) return SQLITE_NOMEM;
1988- memset(pConfig, 0, sizeof(*pConfig));
1989- rc = sqlite3_create_function_v2(db, "editdist3",
1990- 2, SQLITE_UTF8|SQLITE_DETERMINISTIC, pConfig,
1991- editDist3SqlFunc, 0, 0, 0);
1992- if( rc==SQLITE_OK ){
1993+static int editDist3Install(sqlite3 *db)
1994+{
1995+ int rc;
1996+ EditDist3Config *pConfig = sqlite3_malloc64( sizeof(*pConfig) );
1997+ if( pConfig==0 ) return SQLITE_NOMEM;
1998+ memset(pConfig, 0, sizeof(*pConfig));
1999 rc = sqlite3_create_function_v2(db, "editdist3",
2000- 3, SQLITE_UTF8|SQLITE_DETERMINISTIC, pConfig,
2001- editDist3SqlFunc, 0, 0, 0);
2002- }
2003- if( rc==SQLITE_OK ){
2004- rc = sqlite3_create_function_v2(db, "editdist3",
2005- 1, SQLITE_UTF8|SQLITE_DETERMINISTIC, pConfig,
2006- editDist3SqlFunc, 0, 0, editDist3ConfigDelete);
2007- }else{
2008- sqlite3_free(pConfig);
2009- }
2010- return rc;
2011+ 2, SQLITE_UTF8|SQLITE_DETERMINISTIC, pConfig,
2012+ editDist3SqlFunc, 0, 0, 0);
2013+ if( rc==SQLITE_OK ) {
2014+ rc = sqlite3_create_function_v2(db, "editdist3",
2015+ 3, SQLITE_UTF8|SQLITE_DETERMINISTIC, pConfig,
2016+ editDist3SqlFunc, 0, 0, 0);
2017+ }
2018+ if( rc==SQLITE_OK ) {
2019+ rc = sqlite3_create_function_v2(db, "editdist3",
2020+ 1, SQLITE_UTF8|SQLITE_DETERMINISTIC, pConfig,
2021+ editDist3SqlFunc, 0, 0, editDist3ConfigDelete);
2022+ } else {
2023+ sqlite3_free(pConfig);
2024+ }
2025+ return rc;
2026 }
2027 /* End configurable cost unicode edit distance routines
2028 ******************************************************************************
2029@@ -1241,62 +1277,64 @@ ** This lookup table is used to help decode the first byte of
2030 ** a multi-byte UTF8 character.
2031 */
2032 static const unsigned char sqlite3Utf8Trans1[] = {
2033- 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
2034- 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
2035- 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17,
2036- 0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f,
2037- 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
2038- 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
2039- 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
2040- 0x00, 0x01, 0x02, 0x03, 0x00, 0x01, 0x00, 0x00,
2041+ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
2042+ 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
2043+ 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17,
2044+ 0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f,
2045+ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
2046+ 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
2047+ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
2048+ 0x00, 0x01, 0x02, 0x03, 0x00, 0x01, 0x00, 0x00,
2049 };
2050 #endif
2051
2052 /*
2053 ** Return the value of the first UTF-8 character in the string.
2054 */
2055-static int utf8Read(const unsigned char *z, int n, int *pSize){
2056- int c, i;
2057+static int utf8Read(const unsigned char *z, int n, int *pSize)
2058+{
2059+ int c, i;
2060
2061- /* All callers to this routine (in the current implementation)
2062- ** always have n>0. */
2063- if( NEVER(n==0) ){
2064- c = i = 0;
2065- }else{
2066- c = z[0];
2067- i = 1;
2068- if( c>=0xc0 ){
2069- c = sqlite3Utf8Trans1[c-0xc0];
2070- while( i<n && (z[i] & 0xc0)==0x80 ){
2071- c = (c<<6) + (0x3f & z[i++]);
2072- }
2073+ /* All callers to this routine (in the current implementation)
2074+ ** always have n>0. */
2075+ if( NEVER(n==0) ) {
2076+ c = i = 0;
2077+ } else {
2078+ c = z[0];
2079+ i = 1;
2080+ if( c>=0xc0 ) {
2081+ c = sqlite3Utf8Trans1[c-0xc0];
2082+ while( i<n && (z[i] & 0xc0)==0x80 ) {
2083+ c = (c<<6) + (0x3f & z[i++]);
2084+ }
2085+ }
2086 }
2087- }
2088- *pSize = i;
2089- return c;
2090+ *pSize = i;
2091+ return c;
2092 }
2093
2094 /*
2095 ** Return the number of characters in the utf-8 string in the nIn byte
2096 ** buffer pointed to by zIn.
2097 */
2098-static int utf8Charlen(const char *zIn, int nIn){
2099- int i;
2100- int nChar = 0;
2101- for(i=0; i<nIn; nChar++){
2102- int sz;
2103- utf8Read((const unsigned char *)&zIn[i], nIn-i, &sz);
2104- i += sz;
2105- }
2106- return nChar;
2107+static int utf8Charlen(const char *zIn, int nIn)
2108+{
2109+ int i;
2110+ int nChar = 0;
2111+ for(i=0; i<nIn; nChar++) {
2112+ int sz;
2113+ utf8Read((const unsigned char *)&zIn[i], nIn-i, &sz);
2114+ i += sz;
2115+ }
2116+ return nChar;
2117 }
2118
2119 typedef struct Transliteration Transliteration;
2120 struct Transliteration {
2121- unsigned short int cFrom;
2122- unsigned char cTo0, cTo1, cTo2, cTo3;
2123+ unsigned short int cFrom;
2124+ unsigned char cTo0, cTo1, cTo2, cTo3;
2125 #ifdef SQLITE_SPELLFIX_5BYTE_MAPPINGS
2126- unsigned char cTo4;
2127+ unsigned char cTo4;
2128 #endif
2129 };
2130
2131@@ -1304,400 +1342,401 @@ /*
2132 ** Table of translations from unicode characters into ASCII.
2133 */
2134 static const Transliteration translit[] = {
2135- { 0x00A0, 0x20, 0x00, 0x00, 0x00 }, /* to */
2136- { 0x00B5, 0x75, 0x00, 0x00, 0x00 }, /* µ to u */
2137- { 0x00C0, 0x41, 0x00, 0x00, 0x00 }, /* À to A */
2138- { 0x00C1, 0x41, 0x00, 0x00, 0x00 }, /* Á to A */
2139- { 0x00C2, 0x41, 0x00, 0x00, 0x00 }, /* Â to A */
2140- { 0x00C3, 0x41, 0x00, 0x00, 0x00 }, /* Ã to A */
2141- { 0x00C4, 0x41, 0x65, 0x00, 0x00 }, /* Ä to Ae */
2142- { 0x00C5, 0x41, 0x61, 0x00, 0x00 }, /* Å to Aa */
2143- { 0x00C6, 0x41, 0x45, 0x00, 0x00 }, /* Æ to AE */
2144- { 0x00C7, 0x43, 0x00, 0x00, 0x00 }, /* Ç to C */
2145- { 0x00C8, 0x45, 0x00, 0x00, 0x00 }, /* È to E */
2146- { 0x00C9, 0x45, 0x00, 0x00, 0x00 }, /* É to E */
2147- { 0x00CA, 0x45, 0x00, 0x00, 0x00 }, /* Ê to E */
2148- { 0x00CB, 0x45, 0x00, 0x00, 0x00 }, /* Ë to E */
2149- { 0x00CC, 0x49, 0x00, 0x00, 0x00 }, /* Ì to I */
2150- { 0x00CD, 0x49, 0x00, 0x00, 0x00 }, /* Í to I */
2151- { 0x00CE, 0x49, 0x00, 0x00, 0x00 }, /* Î to I */
2152- { 0x00CF, 0x49, 0x00, 0x00, 0x00 }, /* Ï to I */
2153- { 0x00D0, 0x44, 0x00, 0x00, 0x00 }, /* Ð to D */
2154- { 0x00D1, 0x4E, 0x00, 0x00, 0x00 }, /* Ñ to N */
2155- { 0x00D2, 0x4F, 0x00, 0x00, 0x00 }, /* Ò to O */
2156- { 0x00D3, 0x4F, 0x00, 0x00, 0x00 }, /* Ó to O */
2157- { 0x00D4, 0x4F, 0x00, 0x00, 0x00 }, /* Ô to O */
2158- { 0x00D5, 0x4F, 0x00, 0x00, 0x00 }, /* Õ to O */
2159- { 0x00D6, 0x4F, 0x65, 0x00, 0x00 }, /* Ö to Oe */
2160- { 0x00D7, 0x78, 0x00, 0x00, 0x00 }, /* × to x */
2161- { 0x00D8, 0x4F, 0x00, 0x00, 0x00 }, /* Ø to O */
2162- { 0x00D9, 0x55, 0x00, 0x00, 0x00 }, /* Ù to U */
2163- { 0x00DA, 0x55, 0x00, 0x00, 0x00 }, /* Ú to U */
2164- { 0x00DB, 0x55, 0x00, 0x00, 0x00 }, /* Û to U */
2165- { 0x00DC, 0x55, 0x65, 0x00, 0x00 }, /* Ü to Ue */
2166- { 0x00DD, 0x59, 0x00, 0x00, 0x00 }, /* Ý to Y */
2167- { 0x00DE, 0x54, 0x68, 0x00, 0x00 }, /* Þ to Th */
2168- { 0x00DF, 0x73, 0x73, 0x00, 0x00 }, /* ß to ss */
2169- { 0x00E0, 0x61, 0x00, 0x00, 0x00 }, /* à to a */
2170- { 0x00E1, 0x61, 0x00, 0x00, 0x00 }, /* á to a */
2171- { 0x00E2, 0x61, 0x00, 0x00, 0x00 }, /* â to a */
2172- { 0x00E3, 0x61, 0x00, 0x00, 0x00 }, /* ã to a */
2173- { 0x00E4, 0x61, 0x65, 0x00, 0x00 }, /* ä to ae */
2174- { 0x00E5, 0x61, 0x61, 0x00, 0x00 }, /* å to aa */
2175- { 0x00E6, 0x61, 0x65, 0x00, 0x00 }, /* æ to ae */
2176- { 0x00E7, 0x63, 0x00, 0x00, 0x00 }, /* ç to c */
2177- { 0x00E8, 0x65, 0x00, 0x00, 0x00 }, /* è to e */
2178- { 0x00E9, 0x65, 0x00, 0x00, 0x00 }, /* é to e */
2179- { 0x00EA, 0x65, 0x00, 0x00, 0x00 }, /* ê to e */
2180- { 0x00EB, 0x65, 0x00, 0x00, 0x00 }, /* ë to e */
2181- { 0x00EC, 0x69, 0x00, 0x00, 0x00 }, /* ì to i */
2182- { 0x00ED, 0x69, 0x00, 0x00, 0x00 }, /* í to i */
2183- { 0x00EE, 0x69, 0x00, 0x00, 0x00 }, /* î to i */
2184- { 0x00EF, 0x69, 0x00, 0x00, 0x00 }, /* ï to i */
2185- { 0x00F0, 0x64, 0x00, 0x00, 0x00 }, /* ð to d */
2186- { 0x00F1, 0x6E, 0x00, 0x00, 0x00 }, /* ñ to n */
2187- { 0x00F2, 0x6F, 0x00, 0x00, 0x00 }, /* ò to o */
2188- { 0x00F3, 0x6F, 0x00, 0x00, 0x00 }, /* ó to o */
2189- { 0x00F4, 0x6F, 0x00, 0x00, 0x00 }, /* ô to o */
2190- { 0x00F5, 0x6F, 0x00, 0x00, 0x00 }, /* õ to o */
2191- { 0x00F6, 0x6F, 0x65, 0x00, 0x00 }, /* ö to oe */
2192- { 0x00F7, 0x3A, 0x00, 0x00, 0x00 }, /* ÷ to : */
2193- { 0x00F8, 0x6F, 0x00, 0x00, 0x00 }, /* ø to o */
2194- { 0x00F9, 0x75, 0x00, 0x00, 0x00 }, /* ù to u */
2195- { 0x00FA, 0x75, 0x00, 0x00, 0x00 }, /* ú to u */
2196- { 0x00FB, 0x75, 0x00, 0x00, 0x00 }, /* û to u */
2197- { 0x00FC, 0x75, 0x65, 0x00, 0x00 }, /* ü to ue */
2198- { 0x00FD, 0x79, 0x00, 0x00, 0x00 }, /* ý to y */
2199- { 0x00FE, 0x74, 0x68, 0x00, 0x00 }, /* þ to th */
2200- { 0x00FF, 0x79, 0x00, 0x00, 0x00 }, /* ÿ to y */
2201- { 0x0100, 0x41, 0x00, 0x00, 0x00 }, /* Ā to A */
2202- { 0x0101, 0x61, 0x00, 0x00, 0x00 }, /* ā to a */
2203- { 0x0102, 0x41, 0x00, 0x00, 0x00 }, /* Ă to A */
2204- { 0x0103, 0x61, 0x00, 0x00, 0x00 }, /* ă to a */
2205- { 0x0104, 0x41, 0x00, 0x00, 0x00 }, /* Ą to A */
2206- { 0x0105, 0x61, 0x00, 0x00, 0x00 }, /* ą to a */
2207- { 0x0106, 0x43, 0x00, 0x00, 0x00 }, /* Ć to C */
2208- { 0x0107, 0x63, 0x00, 0x00, 0x00 }, /* ć to c */
2209- { 0x0108, 0x43, 0x68, 0x00, 0x00 }, /* Ĉ to Ch */
2210- { 0x0109, 0x63, 0x68, 0x00, 0x00 }, /* ĉ to ch */
2211- { 0x010A, 0x43, 0x00, 0x00, 0x00 }, /* Ċ to C */
2212- { 0x010B, 0x63, 0x00, 0x00, 0x00 }, /* ċ to c */
2213- { 0x010C, 0x43, 0x00, 0x00, 0x00 }, /* Č to C */
2214- { 0x010D, 0x63, 0x00, 0x00, 0x00 }, /* č to c */
2215- { 0x010E, 0x44, 0x00, 0x00, 0x00 }, /* Ď to D */
2216- { 0x010F, 0x64, 0x00, 0x00, 0x00 }, /* ď to d */
2217- { 0x0110, 0x44, 0x00, 0x00, 0x00 }, /* Đ to D */
2218- { 0x0111, 0x64, 0x00, 0x00, 0x00 }, /* đ to d */
2219- { 0x0112, 0x45, 0x00, 0x00, 0x00 }, /* Ē to E */
2220- { 0x0113, 0x65, 0x00, 0x00, 0x00 }, /* ē to e */
2221- { 0x0114, 0x45, 0x00, 0x00, 0x00 }, /* Ĕ to E */
2222- { 0x0115, 0x65, 0x00, 0x00, 0x00 }, /* ĕ to e */
2223- { 0x0116, 0x45, 0x00, 0x00, 0x00 }, /* Ė to E */
2224- { 0x0117, 0x65, 0x00, 0x00, 0x00 }, /* ė to e */
2225- { 0x0118, 0x45, 0x00, 0x00, 0x00 }, /* Ę to E */
2226- { 0x0119, 0x65, 0x00, 0x00, 0x00 }, /* ę to e */
2227- { 0x011A, 0x45, 0x00, 0x00, 0x00 }, /* Ě to E */
2228- { 0x011B, 0x65, 0x00, 0x00, 0x00 }, /* ě to e */
2229- { 0x011C, 0x47, 0x68, 0x00, 0x00 }, /* Ĝ to Gh */
2230- { 0x011D, 0x67, 0x68, 0x00, 0x00 }, /* ĝ to gh */
2231- { 0x011E, 0x47, 0x00, 0x00, 0x00 }, /* Ğ to G */
2232- { 0x011F, 0x67, 0x00, 0x00, 0x00 }, /* ğ to g */
2233- { 0x0120, 0x47, 0x00, 0x00, 0x00 }, /* Ġ to G */
2234- { 0x0121, 0x67, 0x00, 0x00, 0x00 }, /* ġ to g */
2235- { 0x0122, 0x47, 0x00, 0x00, 0x00 }, /* Ģ to G */
2236- { 0x0123, 0x67, 0x00, 0x00, 0x00 }, /* ģ to g */
2237- { 0x0124, 0x48, 0x68, 0x00, 0x00 }, /* Ĥ to Hh */
2238- { 0x0125, 0x68, 0x68, 0x00, 0x00 }, /* ĥ to hh */
2239- { 0x0126, 0x48, 0x00, 0x00, 0x00 }, /* Ħ to H */
2240- { 0x0127, 0x68, 0x00, 0x00, 0x00 }, /* ħ to h */
2241- { 0x0128, 0x49, 0x00, 0x00, 0x00 }, /* Ĩ to I */
2242- { 0x0129, 0x69, 0x00, 0x00, 0x00 }, /* ĩ to i */
2243- { 0x012A, 0x49, 0x00, 0x00, 0x00 }, /* Ī to I */
2244- { 0x012B, 0x69, 0x00, 0x00, 0x00 }, /* ī to i */
2245- { 0x012C, 0x49, 0x00, 0x00, 0x00 }, /* Ĭ to I */
2246- { 0x012D, 0x69, 0x00, 0x00, 0x00 }, /* ĭ to i */
2247- { 0x012E, 0x49, 0x00, 0x00, 0x00 }, /* Į to I */
2248- { 0x012F, 0x69, 0x00, 0x00, 0x00 }, /* į to i */
2249- { 0x0130, 0x49, 0x00, 0x00, 0x00 }, /* İ to I */
2250- { 0x0131, 0x69, 0x00, 0x00, 0x00 }, /* ı to i */
2251- { 0x0132, 0x49, 0x4A, 0x00, 0x00 }, /* IJ to IJ */
2252- { 0x0133, 0x69, 0x6A, 0x00, 0x00 }, /* ij to ij */
2253- { 0x0134, 0x4A, 0x68, 0x00, 0x00 }, /* Ĵ to Jh */
2254- { 0x0135, 0x6A, 0x68, 0x00, 0x00 }, /* ĵ to jh */
2255- { 0x0136, 0x4B, 0x00, 0x00, 0x00 }, /* Ķ to K */
2256- { 0x0137, 0x6B, 0x00, 0x00, 0x00 }, /* ķ to k */
2257- { 0x0138, 0x6B, 0x00, 0x00, 0x00 }, /* ĸ to k */
2258- { 0x0139, 0x4C, 0x00, 0x00, 0x00 }, /* Ĺ to L */
2259- { 0x013A, 0x6C, 0x00, 0x00, 0x00 }, /* ĺ to l */
2260- { 0x013B, 0x4C, 0x00, 0x00, 0x00 }, /* Ļ to L */
2261- { 0x013C, 0x6C, 0x00, 0x00, 0x00 }, /* ļ to l */
2262- { 0x013D, 0x4C, 0x00, 0x00, 0x00 }, /* Ľ to L */
2263- { 0x013E, 0x6C, 0x00, 0x00, 0x00 }, /* ľ to l */
2264- { 0x013F, 0x4C, 0x2E, 0x00, 0x00 }, /* Ŀ to L. */
2265- { 0x0140, 0x6C, 0x2E, 0x00, 0x00 }, /* ŀ to l. */
2266- { 0x0141, 0x4C, 0x00, 0x00, 0x00 }, /* Ł to L */
2267- { 0x0142, 0x6C, 0x00, 0x00, 0x00 }, /* ł to l */
2268- { 0x0143, 0x4E, 0x00, 0x00, 0x00 }, /* Ń to N */
2269- { 0x0144, 0x6E, 0x00, 0x00, 0x00 }, /* ń to n */
2270- { 0x0145, 0x4E, 0x00, 0x00, 0x00 }, /* Ņ to N */
2271- { 0x0146, 0x6E, 0x00, 0x00, 0x00 }, /* ņ to n */
2272- { 0x0147, 0x4E, 0x00, 0x00, 0x00 }, /* Ň to N */
2273- { 0x0148, 0x6E, 0x00, 0x00, 0x00 }, /* ň to n */
2274- { 0x0149, 0x27, 0x6E, 0x00, 0x00 }, /* ʼn to 'n */
2275- { 0x014A, 0x4E, 0x47, 0x00, 0x00 }, /* Ŋ to NG */
2276- { 0x014B, 0x6E, 0x67, 0x00, 0x00 }, /* ŋ to ng */
2277- { 0x014C, 0x4F, 0x00, 0x00, 0x00 }, /* Ō to O */
2278- { 0x014D, 0x6F, 0x00, 0x00, 0x00 }, /* ō to o */
2279- { 0x014E, 0x4F, 0x00, 0x00, 0x00 }, /* Ŏ to O */
2280- { 0x014F, 0x6F, 0x00, 0x00, 0x00 }, /* ŏ to o */
2281- { 0x0150, 0x4F, 0x00, 0x00, 0x00 }, /* Ő to O */
2282- { 0x0151, 0x6F, 0x00, 0x00, 0x00 }, /* ő to o */
2283- { 0x0152, 0x4F, 0x45, 0x00, 0x00 }, /* Πto OE */
2284- { 0x0153, 0x6F, 0x65, 0x00, 0x00 }, /* œ to oe */
2285- { 0x0154, 0x52, 0x00, 0x00, 0x00 }, /* Ŕ to R */
2286- { 0x0155, 0x72, 0x00, 0x00, 0x00 }, /* ŕ to r */
2287- { 0x0156, 0x52, 0x00, 0x00, 0x00 }, /* Ŗ to R */
2288- { 0x0157, 0x72, 0x00, 0x00, 0x00 }, /* ŗ to r */
2289- { 0x0158, 0x52, 0x00, 0x00, 0x00 }, /* Ř to R */
2290- { 0x0159, 0x72, 0x00, 0x00, 0x00 }, /* ř to r */
2291- { 0x015A, 0x53, 0x00, 0x00, 0x00 }, /* Ś to S */
2292- { 0x015B, 0x73, 0x00, 0x00, 0x00 }, /* ś to s */
2293- { 0x015C, 0x53, 0x68, 0x00, 0x00 }, /* Ŝ to Sh */
2294- { 0x015D, 0x73, 0x68, 0x00, 0x00 }, /* ŝ to sh */
2295- { 0x015E, 0x53, 0x00, 0x00, 0x00 }, /* Ş to S */
2296- { 0x015F, 0x73, 0x00, 0x00, 0x00 }, /* ş to s */
2297- { 0x0160, 0x53, 0x00, 0x00, 0x00 }, /* Š to S */
2298- { 0x0161, 0x73, 0x00, 0x00, 0x00 }, /* š to s */
2299- { 0x0162, 0x54, 0x00, 0x00, 0x00 }, /* Ţ to T */
2300- { 0x0163, 0x74, 0x00, 0x00, 0x00 }, /* ţ to t */
2301- { 0x0164, 0x54, 0x00, 0x00, 0x00 }, /* Ť to T */
2302- { 0x0165, 0x74, 0x00, 0x00, 0x00 }, /* ť to t */
2303- { 0x0166, 0x54, 0x00, 0x00, 0x00 }, /* Ŧ to T */
2304- { 0x0167, 0x74, 0x00, 0x00, 0x00 }, /* ŧ to t */
2305- { 0x0168, 0x55, 0x00, 0x00, 0x00 }, /* Ũ to U */
2306- { 0x0169, 0x75, 0x00, 0x00, 0x00 }, /* ũ to u */
2307- { 0x016A, 0x55, 0x00, 0x00, 0x00 }, /* Ū to U */
2308- { 0x016B, 0x75, 0x00, 0x00, 0x00 }, /* ū to u */
2309- { 0x016C, 0x55, 0x00, 0x00, 0x00 }, /* Ŭ to U */
2310- { 0x016D, 0x75, 0x00, 0x00, 0x00 }, /* ŭ to u */
2311- { 0x016E, 0x55, 0x00, 0x00, 0x00 }, /* Ů to U */
2312- { 0x016F, 0x75, 0x00, 0x00, 0x00 }, /* ů to u */
2313- { 0x0170, 0x55, 0x00, 0x00, 0x00 }, /* Ű to U */
2314- { 0x0171, 0x75, 0x00, 0x00, 0x00 }, /* ű to u */
2315- { 0x0172, 0x55, 0x00, 0x00, 0x00 }, /* Ų to U */
2316- { 0x0173, 0x75, 0x00, 0x00, 0x00 }, /* ų to u */
2317- { 0x0174, 0x57, 0x00, 0x00, 0x00 }, /* Ŵ to W */
2318- { 0x0175, 0x77, 0x00, 0x00, 0x00 }, /* ŵ to w */
2319- { 0x0176, 0x59, 0x00, 0x00, 0x00 }, /* Ŷ to Y */
2320- { 0x0177, 0x79, 0x00, 0x00, 0x00 }, /* ŷ to y */
2321- { 0x0178, 0x59, 0x00, 0x00, 0x00 }, /* Ÿ to Y */
2322- { 0x0179, 0x5A, 0x00, 0x00, 0x00 }, /* Ź to Z */
2323- { 0x017A, 0x7A, 0x00, 0x00, 0x00 }, /* ź to z */
2324- { 0x017B, 0x5A, 0x00, 0x00, 0x00 }, /* Ż to Z */
2325- { 0x017C, 0x7A, 0x00, 0x00, 0x00 }, /* ż to z */
2326- { 0x017D, 0x5A, 0x00, 0x00, 0x00 }, /* Ž to Z */
2327- { 0x017E, 0x7A, 0x00, 0x00, 0x00 }, /* ž to z */
2328- { 0x017F, 0x73, 0x00, 0x00, 0x00 }, /* ſ to s */
2329- { 0x0192, 0x66, 0x00, 0x00, 0x00 }, /* ƒ to f */
2330- { 0x0218, 0x53, 0x00, 0x00, 0x00 }, /* Ș to S */
2331- { 0x0219, 0x73, 0x00, 0x00, 0x00 }, /* ș to s */
2332- { 0x021A, 0x54, 0x00, 0x00, 0x00 }, /* Ț to T */
2333- { 0x021B, 0x74, 0x00, 0x00, 0x00 }, /* ț to t */
2334- { 0x0386, 0x41, 0x00, 0x00, 0x00 }, /* Ά to A */
2335- { 0x0388, 0x45, 0x00, 0x00, 0x00 }, /* Έ to E */
2336- { 0x0389, 0x49, 0x00, 0x00, 0x00 }, /* Ή to I */
2337- { 0x038A, 0x49, 0x00, 0x00, 0x00 }, /* Ί to I */
2338- { 0x038C, 0x4f, 0x00, 0x00, 0x00 }, /* Ό to O */
2339- { 0x038E, 0x59, 0x00, 0x00, 0x00 }, /* Ύ to Y */
2340- { 0x038F, 0x4f, 0x00, 0x00, 0x00 }, /* Ώ to O */
2341- { 0x0390, 0x69, 0x00, 0x00, 0x00 }, /* ΐ to i */
2342- { 0x0391, 0x41, 0x00, 0x00, 0x00 }, /* Α to A */
2343- { 0x0392, 0x42, 0x00, 0x00, 0x00 }, /* Β to B */
2344- { 0x0393, 0x47, 0x00, 0x00, 0x00 }, /* Γ to G */
2345- { 0x0394, 0x44, 0x00, 0x00, 0x00 }, /* Δ to D */
2346- { 0x0395, 0x45, 0x00, 0x00, 0x00 }, /* Ε to E */
2347- { 0x0396, 0x5a, 0x00, 0x00, 0x00 }, /* Ζ to Z */
2348- { 0x0397, 0x49, 0x00, 0x00, 0x00 }, /* Η to I */
2349- { 0x0398, 0x54, 0x68, 0x00, 0x00 }, /* Θ to Th */
2350- { 0x0399, 0x49, 0x00, 0x00, 0x00 }, /* Ι to I */
2351- { 0x039A, 0x4b, 0x00, 0x00, 0x00 }, /* Κ to K */
2352- { 0x039B, 0x4c, 0x00, 0x00, 0x00 }, /* Λ to L */
2353- { 0x039C, 0x4d, 0x00, 0x00, 0x00 }, /* Μ to M */
2354- { 0x039D, 0x4e, 0x00, 0x00, 0x00 }, /* Ν to N */
2355- { 0x039E, 0x58, 0x00, 0x00, 0x00 }, /* Ξ to X */
2356- { 0x039F, 0x4f, 0x00, 0x00, 0x00 }, /* Ο to O */
2357- { 0x03A0, 0x50, 0x00, 0x00, 0x00 }, /* Π to P */
2358- { 0x03A1, 0x52, 0x00, 0x00, 0x00 }, /* Ρ to R */
2359- { 0x03A3, 0x53, 0x00, 0x00, 0x00 }, /* Σ to S */
2360- { 0x03A4, 0x54, 0x00, 0x00, 0x00 }, /* Τ to T */
2361- { 0x03A5, 0x59, 0x00, 0x00, 0x00 }, /* Υ to Y */
2362- { 0x03A6, 0x46, 0x00, 0x00, 0x00 }, /* Φ to F */
2363- { 0x03A7, 0x43, 0x68, 0x00, 0x00 }, /* Χ to Ch */
2364- { 0x03A8, 0x50, 0x73, 0x00, 0x00 }, /* Ψ to Ps */
2365- { 0x03A9, 0x4f, 0x00, 0x00, 0x00 }, /* Ω to O */
2366- { 0x03AA, 0x49, 0x00, 0x00, 0x00 }, /* Ϊ to I */
2367- { 0x03AB, 0x59, 0x00, 0x00, 0x00 }, /* Ϋ to Y */
2368- { 0x03AC, 0x61, 0x00, 0x00, 0x00 }, /* ά to a */
2369- { 0x03AD, 0x65, 0x00, 0x00, 0x00 }, /* έ to e */
2370- { 0x03AE, 0x69, 0x00, 0x00, 0x00 }, /* ή to i */
2371- { 0x03AF, 0x69, 0x00, 0x00, 0x00 }, /* ί to i */
2372- { 0x03B1, 0x61, 0x00, 0x00, 0x00 }, /* α to a */
2373- { 0x03B2, 0x62, 0x00, 0x00, 0x00 }, /* β to b */
2374- { 0x03B3, 0x67, 0x00, 0x00, 0x00 }, /* γ to g */
2375- { 0x03B4, 0x64, 0x00, 0x00, 0x00 }, /* δ to d */
2376- { 0x03B5, 0x65, 0x00, 0x00, 0x00 }, /* ε to e */
2377- { 0x03B6, 0x7a, 0x00, 0x00, 0x00 }, /* ζ to z */
2378- { 0x03B7, 0x69, 0x00, 0x00, 0x00 }, /* η to i */
2379- { 0x03B8, 0x74, 0x68, 0x00, 0x00 }, /* θ to th */
2380- { 0x03B9, 0x69, 0x00, 0x00, 0x00 }, /* ι to i */
2381- { 0x03BA, 0x6b, 0x00, 0x00, 0x00 }, /* κ to k */
2382- { 0x03BB, 0x6c, 0x00, 0x00, 0x00 }, /* λ to l */
2383- { 0x03BC, 0x6d, 0x00, 0x00, 0x00 }, /* μ to m */
2384- { 0x03BD, 0x6e, 0x00, 0x00, 0x00 }, /* ν to n */
2385- { 0x03BE, 0x78, 0x00, 0x00, 0x00 }, /* ξ to x */
2386- { 0x03BF, 0x6f, 0x00, 0x00, 0x00 }, /* ο to o */
2387- { 0x03C0, 0x70, 0x00, 0x00, 0x00 }, /* π to p */
2388- { 0x03C1, 0x72, 0x00, 0x00, 0x00 }, /* ρ to r */
2389- { 0x03C3, 0x73, 0x00, 0x00, 0x00 }, /* σ to s */
2390- { 0x03C4, 0x74, 0x00, 0x00, 0x00 }, /* τ to t */
2391- { 0x03C5, 0x79, 0x00, 0x00, 0x00 }, /* υ to y */
2392- { 0x03C6, 0x66, 0x00, 0x00, 0x00 }, /* φ to f */
2393- { 0x03C7, 0x63, 0x68, 0x00, 0x00 }, /* χ to ch */
2394- { 0x03C8, 0x70, 0x73, 0x00, 0x00 }, /* ψ to ps */
2395- { 0x03C9, 0x6f, 0x00, 0x00, 0x00 }, /* ω to o */
2396- { 0x03CA, 0x69, 0x00, 0x00, 0x00 }, /* ϊ to i */
2397- { 0x03CB, 0x79, 0x00, 0x00, 0x00 }, /* ϋ to y */
2398- { 0x03CC, 0x6f, 0x00, 0x00, 0x00 }, /* ό to o */
2399- { 0x03CD, 0x79, 0x00, 0x00, 0x00 }, /* ύ to y */
2400- { 0x03CE, 0x69, 0x00, 0x00, 0x00 }, /* ώ to i */
2401- { 0x0400, 0x45, 0x00, 0x00, 0x00 }, /* Ѐ to E */
2402- { 0x0401, 0x45, 0x00, 0x00, 0x00 }, /* Ё to E */
2403- { 0x0402, 0x44, 0x00, 0x00, 0x00 }, /* Ђ to D */
2404- { 0x0403, 0x47, 0x00, 0x00, 0x00 }, /* Ѓ to G */
2405- { 0x0404, 0x45, 0x00, 0x00, 0x00 }, /* Є to E */
2406- { 0x0405, 0x5a, 0x00, 0x00, 0x00 }, /* Ѕ to Z */
2407- { 0x0406, 0x49, 0x00, 0x00, 0x00 }, /* І to I */
2408- { 0x0407, 0x49, 0x00, 0x00, 0x00 }, /* Ї to I */
2409- { 0x0408, 0x4a, 0x00, 0x00, 0x00 }, /* Ј to J */
2410- { 0x0409, 0x49, 0x00, 0x00, 0x00 }, /* Љ to I */
2411- { 0x040A, 0x4e, 0x00, 0x00, 0x00 }, /* Њ to N */
2412- { 0x040B, 0x44, 0x00, 0x00, 0x00 }, /* Ћ to D */
2413- { 0x040C, 0x4b, 0x00, 0x00, 0x00 }, /* Ќ to K */
2414- { 0x040D, 0x49, 0x00, 0x00, 0x00 }, /* Ѝ to I */
2415- { 0x040E, 0x55, 0x00, 0x00, 0x00 }, /* Ў to U */
2416- { 0x040F, 0x44, 0x00, 0x00, 0x00 }, /* Џ to D */
2417- { 0x0410, 0x41, 0x00, 0x00, 0x00 }, /* А to A */
2418- { 0x0411, 0x42, 0x00, 0x00, 0x00 }, /* Б to B */
2419- { 0x0412, 0x56, 0x00, 0x00, 0x00 }, /* В to V */
2420- { 0x0413, 0x47, 0x00, 0x00, 0x00 }, /* Г to G */
2421- { 0x0414, 0x44, 0x00, 0x00, 0x00 }, /* Д to D */
2422- { 0x0415, 0x45, 0x00, 0x00, 0x00 }, /* Е to E */
2423- { 0x0416, 0x5a, 0x68, 0x00, 0x00 }, /* Ж to Zh */
2424- { 0x0417, 0x5a, 0x00, 0x00, 0x00 }, /* З to Z */
2425- { 0x0418, 0x49, 0x00, 0x00, 0x00 }, /* И to I */
2426- { 0x0419, 0x49, 0x00, 0x00, 0x00 }, /* Й to I */
2427- { 0x041A, 0x4b, 0x00, 0x00, 0x00 }, /* К to K */
2428- { 0x041B, 0x4c, 0x00, 0x00, 0x00 }, /* Л to L */
2429- { 0x041C, 0x4d, 0x00, 0x00, 0x00 }, /* М to M */
2430- { 0x041D, 0x4e, 0x00, 0x00, 0x00 }, /* Н to N */
2431- { 0x041E, 0x4f, 0x00, 0x00, 0x00 }, /* О to O */
2432- { 0x041F, 0x50, 0x00, 0x00, 0x00 }, /* П to P */
2433- { 0x0420, 0x52, 0x00, 0x00, 0x00 }, /* Р to R */
2434- { 0x0421, 0x53, 0x00, 0x00, 0x00 }, /* С to S */
2435- { 0x0422, 0x54, 0x00, 0x00, 0x00 }, /* Т to T */
2436- { 0x0423, 0x55, 0x00, 0x00, 0x00 }, /* У to U */
2437- { 0x0424, 0x46, 0x00, 0x00, 0x00 }, /* Ф to F */
2438- { 0x0425, 0x4b, 0x68, 0x00, 0x00 }, /* Х to Kh */
2439- { 0x0426, 0x54, 0x63, 0x00, 0x00 }, /* Ц to Tc */
2440- { 0x0427, 0x43, 0x68, 0x00, 0x00 }, /* Ч to Ch */
2441- { 0x0428, 0x53, 0x68, 0x00, 0x00 }, /* Ш to Sh */
2442- { 0x0429, 0x53, 0x68, 0x63, 0x68 }, /* Щ to Shch */
2443- { 0x042A, 0x61, 0x00, 0x00, 0x00 }, /* to A */
2444- { 0x042B, 0x59, 0x00, 0x00, 0x00 }, /* Ы to Y */
2445- { 0x042C, 0x59, 0x00, 0x00, 0x00 }, /* to Y */
2446- { 0x042D, 0x45, 0x00, 0x00, 0x00 }, /* Э to E */
2447- { 0x042E, 0x49, 0x75, 0x00, 0x00 }, /* Ю to Iu */
2448- { 0x042F, 0x49, 0x61, 0x00, 0x00 }, /* Я to Ia */
2449- { 0x0430, 0x61, 0x00, 0x00, 0x00 }, /* а to a */
2450- { 0x0431, 0x62, 0x00, 0x00, 0x00 }, /* б to b */
2451- { 0x0432, 0x76, 0x00, 0x00, 0x00 }, /* в to v */
2452- { 0x0433, 0x67, 0x00, 0x00, 0x00 }, /* г to g */
2453- { 0x0434, 0x64, 0x00, 0x00, 0x00 }, /* д to d */
2454- { 0x0435, 0x65, 0x00, 0x00, 0x00 }, /* е to e */
2455- { 0x0436, 0x7a, 0x68, 0x00, 0x00 }, /* ж to zh */
2456- { 0x0437, 0x7a, 0x00, 0x00, 0x00 }, /* з to z */
2457- { 0x0438, 0x69, 0x00, 0x00, 0x00 }, /* и to i */
2458- { 0x0439, 0x69, 0x00, 0x00, 0x00 }, /* й to i */
2459- { 0x043A, 0x6b, 0x00, 0x00, 0x00 }, /* к to k */
2460- { 0x043B, 0x6c, 0x00, 0x00, 0x00 }, /* л to l */
2461- { 0x043C, 0x6d, 0x00, 0x00, 0x00 }, /* м to m */
2462- { 0x043D, 0x6e, 0x00, 0x00, 0x00 }, /* н to n */
2463- { 0x043E, 0x6f, 0x00, 0x00, 0x00 }, /* о to o */
2464- { 0x043F, 0x70, 0x00, 0x00, 0x00 }, /* п to p */
2465- { 0x0440, 0x72, 0x00, 0x00, 0x00 }, /* р to r */
2466- { 0x0441, 0x73, 0x00, 0x00, 0x00 }, /* с to s */
2467- { 0x0442, 0x74, 0x00, 0x00, 0x00 }, /* т to t */
2468- { 0x0443, 0x75, 0x00, 0x00, 0x00 }, /* у to u */
2469- { 0x0444, 0x66, 0x00, 0x00, 0x00 }, /* ф to f */
2470- { 0x0445, 0x6b, 0x68, 0x00, 0x00 }, /* х to kh */
2471- { 0x0446, 0x74, 0x63, 0x00, 0x00 }, /* ц to tc */
2472- { 0x0447, 0x63, 0x68, 0x00, 0x00 }, /* ч to ch */
2473- { 0x0448, 0x73, 0x68, 0x00, 0x00 }, /* ш to sh */
2474- { 0x0449, 0x73, 0x68, 0x63, 0x68 }, /* щ to shch */
2475- { 0x044A, 0x61, 0x00, 0x00, 0x00 }, /* to a */
2476- { 0x044B, 0x79, 0x00, 0x00, 0x00 }, /* ы to y */
2477- { 0x044C, 0x79, 0x00, 0x00, 0x00 }, /* to y */
2478- { 0x044D, 0x65, 0x00, 0x00, 0x00 }, /* э to e */
2479- { 0x044E, 0x69, 0x75, 0x00, 0x00 }, /* ю to iu */
2480- { 0x044F, 0x69, 0x61, 0x00, 0x00 }, /* я to ia */
2481- { 0x0450, 0x65, 0x00, 0x00, 0x00 }, /* ѐ to e */
2482- { 0x0451, 0x65, 0x00, 0x00, 0x00 }, /* ё to e */
2483- { 0x0452, 0x64, 0x00, 0x00, 0x00 }, /* ђ to d */
2484- { 0x0453, 0x67, 0x00, 0x00, 0x00 }, /* ѓ to g */
2485- { 0x0454, 0x65, 0x00, 0x00, 0x00 }, /* є to e */
2486- { 0x0455, 0x7a, 0x00, 0x00, 0x00 }, /* ѕ to z */
2487- { 0x0456, 0x69, 0x00, 0x00, 0x00 }, /* і to i */
2488- { 0x0457, 0x69, 0x00, 0x00, 0x00 }, /* ї to i */
2489- { 0x0458, 0x6a, 0x00, 0x00, 0x00 }, /* ј to j */
2490- { 0x0459, 0x69, 0x00, 0x00, 0x00 }, /* љ to i */
2491- { 0x045A, 0x6e, 0x00, 0x00, 0x00 }, /* њ to n */
2492- { 0x045B, 0x64, 0x00, 0x00, 0x00 }, /* ћ to d */
2493- { 0x045C, 0x6b, 0x00, 0x00, 0x00 }, /* ќ to k */
2494- { 0x045D, 0x69, 0x00, 0x00, 0x00 }, /* ѝ to i */
2495- { 0x045E, 0x75, 0x00, 0x00, 0x00 }, /* ў to u */
2496- { 0x045F, 0x64, 0x00, 0x00, 0x00 }, /* џ to d */
2497- { 0x1E02, 0x42, 0x00, 0x00, 0x00 }, /* Ḃ to B */
2498- { 0x1E03, 0x62, 0x00, 0x00, 0x00 }, /* ḃ to b */
2499- { 0x1E0A, 0x44, 0x00, 0x00, 0x00 }, /* Ḋ to D */
2500- { 0x1E0B, 0x64, 0x00, 0x00, 0x00 }, /* ḋ to d */
2501- { 0x1E1E, 0x46, 0x00, 0x00, 0x00 }, /* Ḟ to F */
2502- { 0x1E1F, 0x66, 0x00, 0x00, 0x00 }, /* ḟ to f */
2503- { 0x1E40, 0x4D, 0x00, 0x00, 0x00 }, /* Ṁ to M */
2504- { 0x1E41, 0x6D, 0x00, 0x00, 0x00 }, /* ṁ to m */
2505- { 0x1E56, 0x50, 0x00, 0x00, 0x00 }, /* Ṗ to P */
2506- { 0x1E57, 0x70, 0x00, 0x00, 0x00 }, /* ṗ to p */
2507- { 0x1E60, 0x53, 0x00, 0x00, 0x00 }, /* Ṡ to S */
2508- { 0x1E61, 0x73, 0x00, 0x00, 0x00 }, /* ṡ to s */
2509- { 0x1E6A, 0x54, 0x00, 0x00, 0x00 }, /* Ṫ to T */
2510- { 0x1E6B, 0x74, 0x00, 0x00, 0x00 }, /* ṫ to t */
2511- { 0x1E80, 0x57, 0x00, 0x00, 0x00 }, /* Ẁ to W */
2512- { 0x1E81, 0x77, 0x00, 0x00, 0x00 }, /* ẁ to w */
2513- { 0x1E82, 0x57, 0x00, 0x00, 0x00 }, /* Ẃ to W */
2514- { 0x1E83, 0x77, 0x00, 0x00, 0x00 }, /* ẃ to w */
2515- { 0x1E84, 0x57, 0x00, 0x00, 0x00 }, /* Ẅ to W */
2516- { 0x1E85, 0x77, 0x00, 0x00, 0x00 }, /* ẅ to w */
2517- { 0x1EF2, 0x59, 0x00, 0x00, 0x00 }, /* Ỳ to Y */
2518- { 0x1EF3, 0x79, 0x00, 0x00, 0x00 }, /* ỳ to y */
2519- { 0xFB00, 0x66, 0x66, 0x00, 0x00 }, /* ff to ff */
2520- { 0xFB01, 0x66, 0x69, 0x00, 0x00 }, /* fi to fi */
2521- { 0xFB02, 0x66, 0x6C, 0x00, 0x00 }, /* fl to fl */
2522- { 0xFB05, 0x73, 0x74, 0x00, 0x00 }, /* ſt to st */
2523- { 0xFB06, 0x73, 0x74, 0x00, 0x00 }, /* st to st */
2524+ { 0x00A0, 0x20, 0x00, 0x00, 0x00 }, /* to */
2525+ { 0x00B5, 0x75, 0x00, 0x00, 0x00 }, /* µ to u */
2526+ { 0x00C0, 0x41, 0x00, 0x00, 0x00 }, /* À to A */
2527+ { 0x00C1, 0x41, 0x00, 0x00, 0x00 }, /* Á to A */
2528+ { 0x00C2, 0x41, 0x00, 0x00, 0x00 }, /* Â to A */
2529+ { 0x00C3, 0x41, 0x00, 0x00, 0x00 }, /* Ã to A */
2530+ { 0x00C4, 0x41, 0x65, 0x00, 0x00 }, /* Ä to Ae */
2531+ { 0x00C5, 0x41, 0x61, 0x00, 0x00 }, /* Å to Aa */
2532+ { 0x00C6, 0x41, 0x45, 0x00, 0x00 }, /* Æ to AE */
2533+ { 0x00C7, 0x43, 0x00, 0x00, 0x00 }, /* Ç to C */
2534+ { 0x00C8, 0x45, 0x00, 0x00, 0x00 }, /* È to E */
2535+ { 0x00C9, 0x45, 0x00, 0x00, 0x00 }, /* É to E */
2536+ { 0x00CA, 0x45, 0x00, 0x00, 0x00 }, /* Ê to E */
2537+ { 0x00CB, 0x45, 0x00, 0x00, 0x00 }, /* Ë to E */
2538+ { 0x00CC, 0x49, 0x00, 0x00, 0x00 }, /* Ì to I */
2539+ { 0x00CD, 0x49, 0x00, 0x00, 0x00 }, /* Í to I */
2540+ { 0x00CE, 0x49, 0x00, 0x00, 0x00 }, /* Î to I */
2541+ { 0x00CF, 0x49, 0x00, 0x00, 0x00 }, /* Ï to I */
2542+ { 0x00D0, 0x44, 0x00, 0x00, 0x00 }, /* Ð to D */
2543+ { 0x00D1, 0x4E, 0x00, 0x00, 0x00 }, /* Ñ to N */
2544+ { 0x00D2, 0x4F, 0x00, 0x00, 0x00 }, /* Ò to O */
2545+ { 0x00D3, 0x4F, 0x00, 0x00, 0x00 }, /* Ó to O */
2546+ { 0x00D4, 0x4F, 0x00, 0x00, 0x00 }, /* Ô to O */
2547+ { 0x00D5, 0x4F, 0x00, 0x00, 0x00 }, /* Õ to O */
2548+ { 0x00D6, 0x4F, 0x65, 0x00, 0x00 }, /* Ö to Oe */
2549+ { 0x00D7, 0x78, 0x00, 0x00, 0x00 }, /* × to x */
2550+ { 0x00D8, 0x4F, 0x00, 0x00, 0x00 }, /* Ø to O */
2551+ { 0x00D9, 0x55, 0x00, 0x00, 0x00 }, /* Ù to U */
2552+ { 0x00DA, 0x55, 0x00, 0x00, 0x00 }, /* Ú to U */
2553+ { 0x00DB, 0x55, 0x00, 0x00, 0x00 }, /* Û to U */
2554+ { 0x00DC, 0x55, 0x65, 0x00, 0x00 }, /* Ü to Ue */
2555+ { 0x00DD, 0x59, 0x00, 0x00, 0x00 }, /* Ý to Y */
2556+ { 0x00DE, 0x54, 0x68, 0x00, 0x00 }, /* Þ to Th */
2557+ { 0x00DF, 0x73, 0x73, 0x00, 0x00 }, /* ß to ss */
2558+ { 0x00E0, 0x61, 0x00, 0x00, 0x00 }, /* à to a */
2559+ { 0x00E1, 0x61, 0x00, 0x00, 0x00 }, /* á to a */
2560+ { 0x00E2, 0x61, 0x00, 0x00, 0x00 }, /* â to a */
2561+ { 0x00E3, 0x61, 0x00, 0x00, 0x00 }, /* ã to a */
2562+ { 0x00E4, 0x61, 0x65, 0x00, 0x00 }, /* ä to ae */
2563+ { 0x00E5, 0x61, 0x61, 0x00, 0x00 }, /* å to aa */
2564+ { 0x00E6, 0x61, 0x65, 0x00, 0x00 }, /* æ to ae */
2565+ { 0x00E7, 0x63, 0x00, 0x00, 0x00 }, /* ç to c */
2566+ { 0x00E8, 0x65, 0x00, 0x00, 0x00 }, /* è to e */
2567+ { 0x00E9, 0x65, 0x00, 0x00, 0x00 }, /* é to e */
2568+ { 0x00EA, 0x65, 0x00, 0x00, 0x00 }, /* ê to e */
2569+ { 0x00EB, 0x65, 0x00, 0x00, 0x00 }, /* ë to e */
2570+ { 0x00EC, 0x69, 0x00, 0x00, 0x00 }, /* ì to i */
2571+ { 0x00ED, 0x69, 0x00, 0x00, 0x00 }, /* í to i */
2572+ { 0x00EE, 0x69, 0x00, 0x00, 0x00 }, /* î to i */
2573+ { 0x00EF, 0x69, 0x00, 0x00, 0x00 }, /* ï to i */
2574+ { 0x00F0, 0x64, 0x00, 0x00, 0x00 }, /* ð to d */
2575+ { 0x00F1, 0x6E, 0x00, 0x00, 0x00 }, /* ñ to n */
2576+ { 0x00F2, 0x6F, 0x00, 0x00, 0x00 }, /* ò to o */
2577+ { 0x00F3, 0x6F, 0x00, 0x00, 0x00 }, /* ó to o */
2578+ { 0x00F4, 0x6F, 0x00, 0x00, 0x00 }, /* ô to o */
2579+ { 0x00F5, 0x6F, 0x00, 0x00, 0x00 }, /* õ to o */
2580+ { 0x00F6, 0x6F, 0x65, 0x00, 0x00 }, /* ö to oe */
2581+ { 0x00F7, 0x3A, 0x00, 0x00, 0x00 }, /* ÷ to : */
2582+ { 0x00F8, 0x6F, 0x00, 0x00, 0x00 }, /* ø to o */
2583+ { 0x00F9, 0x75, 0x00, 0x00, 0x00 }, /* ù to u */
2584+ { 0x00FA, 0x75, 0x00, 0x00, 0x00 }, /* ú to u */
2585+ { 0x00FB, 0x75, 0x00, 0x00, 0x00 }, /* û to u */
2586+ { 0x00FC, 0x75, 0x65, 0x00, 0x00 }, /* ü to ue */
2587+ { 0x00FD, 0x79, 0x00, 0x00, 0x00 }, /* ý to y */
2588+ { 0x00FE, 0x74, 0x68, 0x00, 0x00 }, /* þ to th */
2589+ { 0x00FF, 0x79, 0x00, 0x00, 0x00 }, /* ÿ to y */
2590+ { 0x0100, 0x41, 0x00, 0x00, 0x00 }, /* Ā to A */
2591+ { 0x0101, 0x61, 0x00, 0x00, 0x00 }, /* ā to a */
2592+ { 0x0102, 0x41, 0x00, 0x00, 0x00 }, /* Ă to A */
2593+ { 0x0103, 0x61, 0x00, 0x00, 0x00 }, /* ă to a */
2594+ { 0x0104, 0x41, 0x00, 0x00, 0x00 }, /* Ą to A */
2595+ { 0x0105, 0x61, 0x00, 0x00, 0x00 }, /* ą to a */
2596+ { 0x0106, 0x43, 0x00, 0x00, 0x00 }, /* Ć to C */
2597+ { 0x0107, 0x63, 0x00, 0x00, 0x00 }, /* ć to c */
2598+ { 0x0108, 0x43, 0x68, 0x00, 0x00 }, /* Ĉ to Ch */
2599+ { 0x0109, 0x63, 0x68, 0x00, 0x00 }, /* ĉ to ch */
2600+ { 0x010A, 0x43, 0x00, 0x00, 0x00 }, /* Ċ to C */
2601+ { 0x010B, 0x63, 0x00, 0x00, 0x00 }, /* ċ to c */
2602+ { 0x010C, 0x43, 0x00, 0x00, 0x00 }, /* Č to C */
2603+ { 0x010D, 0x63, 0x00, 0x00, 0x00 }, /* č to c */
2604+ { 0x010E, 0x44, 0x00, 0x00, 0x00 }, /* Ď to D */
2605+ { 0x010F, 0x64, 0x00, 0x00, 0x00 }, /* ď to d */
2606+ { 0x0110, 0x44, 0x00, 0x00, 0x00 }, /* Đ to D */
2607+ { 0x0111, 0x64, 0x00, 0x00, 0x00 }, /* đ to d */
2608+ { 0x0112, 0x45, 0x00, 0x00, 0x00 }, /* Ē to E */
2609+ { 0x0113, 0x65, 0x00, 0x00, 0x00 }, /* ē to e */
2610+ { 0x0114, 0x45, 0x00, 0x00, 0x00 }, /* Ĕ to E */
2611+ { 0x0115, 0x65, 0x00, 0x00, 0x00 }, /* ĕ to e */
2612+ { 0x0116, 0x45, 0x00, 0x00, 0x00 }, /* Ė to E */
2613+ { 0x0117, 0x65, 0x00, 0x00, 0x00 }, /* ė to e */
2614+ { 0x0118, 0x45, 0x00, 0x00, 0x00 }, /* Ę to E */
2615+ { 0x0119, 0x65, 0x00, 0x00, 0x00 }, /* ę to e */
2616+ { 0x011A, 0x45, 0x00, 0x00, 0x00 }, /* Ě to E */
2617+ { 0x011B, 0x65, 0x00, 0x00, 0x00 }, /* ě to e */
2618+ { 0x011C, 0x47, 0x68, 0x00, 0x00 }, /* Ĝ to Gh */
2619+ { 0x011D, 0x67, 0x68, 0x00, 0x00 }, /* ĝ to gh */
2620+ { 0x011E, 0x47, 0x00, 0x00, 0x00 }, /* Ğ to G */
2621+ { 0x011F, 0x67, 0x00, 0x00, 0x00 }, /* ğ to g */
2622+ { 0x0120, 0x47, 0x00, 0x00, 0x00 }, /* Ġ to G */
2623+ { 0x0121, 0x67, 0x00, 0x00, 0x00 }, /* ġ to g */
2624+ { 0x0122, 0x47, 0x00, 0x00, 0x00 }, /* Ģ to G */
2625+ { 0x0123, 0x67, 0x00, 0x00, 0x00 }, /* ģ to g */
2626+ { 0x0124, 0x48, 0x68, 0x00, 0x00 }, /* Ĥ to Hh */
2627+ { 0x0125, 0x68, 0x68, 0x00, 0x00 }, /* ĥ to hh */
2628+ { 0x0126, 0x48, 0x00, 0x00, 0x00 }, /* Ħ to H */
2629+ { 0x0127, 0x68, 0x00, 0x00, 0x00 }, /* ħ to h */
2630+ { 0x0128, 0x49, 0x00, 0x00, 0x00 }, /* Ĩ to I */
2631+ { 0x0129, 0x69, 0x00, 0x00, 0x00 }, /* ĩ to i */
2632+ { 0x012A, 0x49, 0x00, 0x00, 0x00 }, /* Ī to I */
2633+ { 0x012B, 0x69, 0x00, 0x00, 0x00 }, /* ī to i */
2634+ { 0x012C, 0x49, 0x00, 0x00, 0x00 }, /* Ĭ to I */
2635+ { 0x012D, 0x69, 0x00, 0x00, 0x00 }, /* ĭ to i */
2636+ { 0x012E, 0x49, 0x00, 0x00, 0x00 }, /* Į to I */
2637+ { 0x012F, 0x69, 0x00, 0x00, 0x00 }, /* į to i */
2638+ { 0x0130, 0x49, 0x00, 0x00, 0x00 }, /* İ to I */
2639+ { 0x0131, 0x69, 0x00, 0x00, 0x00 }, /* ı to i */
2640+ { 0x0132, 0x49, 0x4A, 0x00, 0x00 }, /* IJ to IJ */
2641+ { 0x0133, 0x69, 0x6A, 0x00, 0x00 }, /* ij to ij */
2642+ { 0x0134, 0x4A, 0x68, 0x00, 0x00 }, /* Ĵ to Jh */
2643+ { 0x0135, 0x6A, 0x68, 0x00, 0x00 }, /* ĵ to jh */
2644+ { 0x0136, 0x4B, 0x00, 0x00, 0x00 }, /* Ķ to K */
2645+ { 0x0137, 0x6B, 0x00, 0x00, 0x00 }, /* ķ to k */
2646+ { 0x0138, 0x6B, 0x00, 0x00, 0x00 }, /* ĸ to k */
2647+ { 0x0139, 0x4C, 0x00, 0x00, 0x00 }, /* Ĺ to L */
2648+ { 0x013A, 0x6C, 0x00, 0x00, 0x00 }, /* ĺ to l */
2649+ { 0x013B, 0x4C, 0x00, 0x00, 0x00 }, /* Ļ to L */
2650+ { 0x013C, 0x6C, 0x00, 0x00, 0x00 }, /* ļ to l */
2651+ { 0x013D, 0x4C, 0x00, 0x00, 0x00 }, /* Ľ to L */
2652+ { 0x013E, 0x6C, 0x00, 0x00, 0x00 }, /* ľ to l */
2653+ { 0x013F, 0x4C, 0x2E, 0x00, 0x00 }, /* Ŀ to L. */
2654+ { 0x0140, 0x6C, 0x2E, 0x00, 0x00 }, /* ŀ to l. */
2655+ { 0x0141, 0x4C, 0x00, 0x00, 0x00 }, /* Ł to L */
2656+ { 0x0142, 0x6C, 0x00, 0x00, 0x00 }, /* ł to l */
2657+ { 0x0143, 0x4E, 0x00, 0x00, 0x00 }, /* Ń to N */
2658+ { 0x0144, 0x6E, 0x00, 0x00, 0x00 }, /* ń to n */
2659+ { 0x0145, 0x4E, 0x00, 0x00, 0x00 }, /* Ņ to N */
2660+ { 0x0146, 0x6E, 0x00, 0x00, 0x00 }, /* ņ to n */
2661+ { 0x0147, 0x4E, 0x00, 0x00, 0x00 }, /* Ň to N */
2662+ { 0x0148, 0x6E, 0x00, 0x00, 0x00 }, /* ň to n */
2663+ { 0x0149, 0x27, 0x6E, 0x00, 0x00 }, /* ʼn to 'n */
2664+ { 0x014A, 0x4E, 0x47, 0x00, 0x00 }, /* Ŋ to NG */
2665+ { 0x014B, 0x6E, 0x67, 0x00, 0x00 }, /* ŋ to ng */
2666+ { 0x014C, 0x4F, 0x00, 0x00, 0x00 }, /* Ō to O */
2667+ { 0x014D, 0x6F, 0x00, 0x00, 0x00 }, /* ō to o */
2668+ { 0x014E, 0x4F, 0x00, 0x00, 0x00 }, /* Ŏ to O */
2669+ { 0x014F, 0x6F, 0x00, 0x00, 0x00 }, /* ŏ to o */
2670+ { 0x0150, 0x4F, 0x00, 0x00, 0x00 }, /* Ő to O */
2671+ { 0x0151, 0x6F, 0x00, 0x00, 0x00 }, /* ő to o */
2672+ { 0x0152, 0x4F, 0x45, 0x00, 0x00 }, /* Πto OE */
2673+ { 0x0153, 0x6F, 0x65, 0x00, 0x00 }, /* œ to oe */
2674+ { 0x0154, 0x52, 0x00, 0x00, 0x00 }, /* Ŕ to R */
2675+ { 0x0155, 0x72, 0x00, 0x00, 0x00 }, /* ŕ to r */
2676+ { 0x0156, 0x52, 0x00, 0x00, 0x00 }, /* Ŗ to R */
2677+ { 0x0157, 0x72, 0x00, 0x00, 0x00 }, /* ŗ to r */
2678+ { 0x0158, 0x52, 0x00, 0x00, 0x00 }, /* Ř to R */
2679+ { 0x0159, 0x72, 0x00, 0x00, 0x00 }, /* ř to r */
2680+ { 0x015A, 0x53, 0x00, 0x00, 0x00 }, /* Ś to S */
2681+ { 0x015B, 0x73, 0x00, 0x00, 0x00 }, /* ś to s */
2682+ { 0x015C, 0x53, 0x68, 0x00, 0x00 }, /* Ŝ to Sh */
2683+ { 0x015D, 0x73, 0x68, 0x00, 0x00 }, /* ŝ to sh */
2684+ { 0x015E, 0x53, 0x00, 0x00, 0x00 }, /* Ş to S */
2685+ { 0x015F, 0x73, 0x00, 0x00, 0x00 }, /* ş to s */
2686+ { 0x0160, 0x53, 0x00, 0x00, 0x00 }, /* Š to S */
2687+ { 0x0161, 0x73, 0x00, 0x00, 0x00 }, /* š to s */
2688+ { 0x0162, 0x54, 0x00, 0x00, 0x00 }, /* Ţ to T */
2689+ { 0x0163, 0x74, 0x00, 0x00, 0x00 }, /* ţ to t */
2690+ { 0x0164, 0x54, 0x00, 0x00, 0x00 }, /* Ť to T */
2691+ { 0x0165, 0x74, 0x00, 0x00, 0x00 }, /* ť to t */
2692+ { 0x0166, 0x54, 0x00, 0x00, 0x00 }, /* Ŧ to T */
2693+ { 0x0167, 0x74, 0x00, 0x00, 0x00 }, /* ŧ to t */
2694+ { 0x0168, 0x55, 0x00, 0x00, 0x00 }, /* Ũ to U */
2695+ { 0x0169, 0x75, 0x00, 0x00, 0x00 }, /* ũ to u */
2696+ { 0x016A, 0x55, 0x00, 0x00, 0x00 }, /* Ū to U */
2697+ { 0x016B, 0x75, 0x00, 0x00, 0x00 }, /* ū to u */
2698+ { 0x016C, 0x55, 0x00, 0x00, 0x00 }, /* Ŭ to U */
2699+ { 0x016D, 0x75, 0x00, 0x00, 0x00 }, /* ŭ to u */
2700+ { 0x016E, 0x55, 0x00, 0x00, 0x00 }, /* Ů to U */
2701+ { 0x016F, 0x75, 0x00, 0x00, 0x00 }, /* ů to u */
2702+ { 0x0170, 0x55, 0x00, 0x00, 0x00 }, /* Ű to U */
2703+ { 0x0171, 0x75, 0x00, 0x00, 0x00 }, /* ű to u */
2704+ { 0x0172, 0x55, 0x00, 0x00, 0x00 }, /* Ų to U */
2705+ { 0x0173, 0x75, 0x00, 0x00, 0x00 }, /* ų to u */
2706+ { 0x0174, 0x57, 0x00, 0x00, 0x00 }, /* Ŵ to W */
2707+ { 0x0175, 0x77, 0x00, 0x00, 0x00 }, /* ŵ to w */
2708+ { 0x0176, 0x59, 0x00, 0x00, 0x00 }, /* Ŷ to Y */
2709+ { 0x0177, 0x79, 0x00, 0x00, 0x00 }, /* ŷ to y */
2710+ { 0x0178, 0x59, 0x00, 0x00, 0x00 }, /* Ÿ to Y */
2711+ { 0x0179, 0x5A, 0x00, 0x00, 0x00 }, /* Ź to Z */
2712+ { 0x017A, 0x7A, 0x00, 0x00, 0x00 }, /* ź to z */
2713+ { 0x017B, 0x5A, 0x00, 0x00, 0x00 }, /* Ż to Z */
2714+ { 0x017C, 0x7A, 0x00, 0x00, 0x00 }, /* ż to z */
2715+ { 0x017D, 0x5A, 0x00, 0x00, 0x00 }, /* Ž to Z */
2716+ { 0x017E, 0x7A, 0x00, 0x00, 0x00 }, /* ž to z */
2717+ { 0x017F, 0x73, 0x00, 0x00, 0x00 }, /* ſ to s */
2718+ { 0x0192, 0x66, 0x00, 0x00, 0x00 }, /* ƒ to f */
2719+ { 0x0218, 0x53, 0x00, 0x00, 0x00 }, /* Ș to S */
2720+ { 0x0219, 0x73, 0x00, 0x00, 0x00 }, /* ș to s */
2721+ { 0x021A, 0x54, 0x00, 0x00, 0x00 }, /* Ț to T */
2722+ { 0x021B, 0x74, 0x00, 0x00, 0x00 }, /* ț to t */
2723+ { 0x0386, 0x41, 0x00, 0x00, 0x00 }, /* Ά to A */
2724+ { 0x0388, 0x45, 0x00, 0x00, 0x00 }, /* Έ to E */
2725+ { 0x0389, 0x49, 0x00, 0x00, 0x00 }, /* Ή to I */
2726+ { 0x038A, 0x49, 0x00, 0x00, 0x00 }, /* Ί to I */
2727+ { 0x038C, 0x4f, 0x00, 0x00, 0x00 }, /* Ό to O */
2728+ { 0x038E, 0x59, 0x00, 0x00, 0x00 }, /* Ύ to Y */
2729+ { 0x038F, 0x4f, 0x00, 0x00, 0x00 }, /* Ώ to O */
2730+ { 0x0390, 0x69, 0x00, 0x00, 0x00 }, /* ΐ to i */
2731+ { 0x0391, 0x41, 0x00, 0x00, 0x00 }, /* Α to A */
2732+ { 0x0392, 0x42, 0x00, 0x00, 0x00 }, /* Β to B */
2733+ { 0x0393, 0x47, 0x00, 0x00, 0x00 }, /* Γ to G */
2734+ { 0x0394, 0x44, 0x00, 0x00, 0x00 }, /* Δ to D */
2735+ { 0x0395, 0x45, 0x00, 0x00, 0x00 }, /* Ε to E */
2736+ { 0x0396, 0x5a, 0x00, 0x00, 0x00 }, /* Ζ to Z */
2737+ { 0x0397, 0x49, 0x00, 0x00, 0x00 }, /* Η to I */
2738+ { 0x0398, 0x54, 0x68, 0x00, 0x00 }, /* Θ to Th */
2739+ { 0x0399, 0x49, 0x00, 0x00, 0x00 }, /* Ι to I */
2740+ { 0x039A, 0x4b, 0x00, 0x00, 0x00 }, /* Κ to K */
2741+ { 0x039B, 0x4c, 0x00, 0x00, 0x00 }, /* Λ to L */
2742+ { 0x039C, 0x4d, 0x00, 0x00, 0x00 }, /* Μ to M */
2743+ { 0x039D, 0x4e, 0x00, 0x00, 0x00 }, /* Ν to N */
2744+ { 0x039E, 0x58, 0x00, 0x00, 0x00 }, /* Ξ to X */
2745+ { 0x039F, 0x4f, 0x00, 0x00, 0x00 }, /* Ο to O */
2746+ { 0x03A0, 0x50, 0x00, 0x00, 0x00 }, /* Π to P */
2747+ { 0x03A1, 0x52, 0x00, 0x00, 0x00 }, /* Ρ to R */
2748+ { 0x03A3, 0x53, 0x00, 0x00, 0x00 }, /* Σ to S */
2749+ { 0x03A4, 0x54, 0x00, 0x00, 0x00 }, /* Τ to T */
2750+ { 0x03A5, 0x59, 0x00, 0x00, 0x00 }, /* Υ to Y */
2751+ { 0x03A6, 0x46, 0x00, 0x00, 0x00 }, /* Φ to F */
2752+ { 0x03A7, 0x43, 0x68, 0x00, 0x00 }, /* Χ to Ch */
2753+ { 0x03A8, 0x50, 0x73, 0x00, 0x00 }, /* Ψ to Ps */
2754+ { 0x03A9, 0x4f, 0x00, 0x00, 0x00 }, /* Ω to O */
2755+ { 0x03AA, 0x49, 0x00, 0x00, 0x00 }, /* Ϊ to I */
2756+ { 0x03AB, 0x59, 0x00, 0x00, 0x00 }, /* Ϋ to Y */
2757+ { 0x03AC, 0x61, 0x00, 0x00, 0x00 }, /* ά to a */
2758+ { 0x03AD, 0x65, 0x00, 0x00, 0x00 }, /* έ to e */
2759+ { 0x03AE, 0x69, 0x00, 0x00, 0x00 }, /* ή to i */
2760+ { 0x03AF, 0x69, 0x00, 0x00, 0x00 }, /* ί to i */
2761+ { 0x03B1, 0x61, 0x00, 0x00, 0x00 }, /* α to a */
2762+ { 0x03B2, 0x62, 0x00, 0x00, 0x00 }, /* β to b */
2763+ { 0x03B3, 0x67, 0x00, 0x00, 0x00 }, /* γ to g */
2764+ { 0x03B4, 0x64, 0x00, 0x00, 0x00 }, /* δ to d */
2765+ { 0x03B5, 0x65, 0x00, 0x00, 0x00 }, /* ε to e */
2766+ { 0x03B6, 0x7a, 0x00, 0x00, 0x00 }, /* ζ to z */
2767+ { 0x03B7, 0x69, 0x00, 0x00, 0x00 }, /* η to i */
2768+ { 0x03B8, 0x74, 0x68, 0x00, 0x00 }, /* θ to th */
2769+ { 0x03B9, 0x69, 0x00, 0x00, 0x00 }, /* ι to i */
2770+ { 0x03BA, 0x6b, 0x00, 0x00, 0x00 }, /* κ to k */
2771+ { 0x03BB, 0x6c, 0x00, 0x00, 0x00 }, /* λ to l */
2772+ { 0x03BC, 0x6d, 0x00, 0x00, 0x00 }, /* μ to m */
2773+ { 0x03BD, 0x6e, 0x00, 0x00, 0x00 }, /* ν to n */
2774+ { 0x03BE, 0x78, 0x00, 0x00, 0x00 }, /* ξ to x */
2775+ { 0x03BF, 0x6f, 0x00, 0x00, 0x00 }, /* ο to o */
2776+ { 0x03C0, 0x70, 0x00, 0x00, 0x00 }, /* π to p */
2777+ { 0x03C1, 0x72, 0x00, 0x00, 0x00 }, /* ρ to r */
2778+ { 0x03C3, 0x73, 0x00, 0x00, 0x00 }, /* σ to s */
2779+ { 0x03C4, 0x74, 0x00, 0x00, 0x00 }, /* τ to t */
2780+ { 0x03C5, 0x79, 0x00, 0x00, 0x00 }, /* υ to y */
2781+ { 0x03C6, 0x66, 0x00, 0x00, 0x00 }, /* φ to f */
2782+ { 0x03C7, 0x63, 0x68, 0x00, 0x00 }, /* χ to ch */
2783+ { 0x03C8, 0x70, 0x73, 0x00, 0x00 }, /* ψ to ps */
2784+ { 0x03C9, 0x6f, 0x00, 0x00, 0x00 }, /* ω to o */
2785+ { 0x03CA, 0x69, 0x00, 0x00, 0x00 }, /* ϊ to i */
2786+ { 0x03CB, 0x79, 0x00, 0x00, 0x00 }, /* ϋ to y */
2787+ { 0x03CC, 0x6f, 0x00, 0x00, 0x00 }, /* ό to o */
2788+ { 0x03CD, 0x79, 0x00, 0x00, 0x00 }, /* ύ to y */
2789+ { 0x03CE, 0x69, 0x00, 0x00, 0x00 }, /* ώ to i */
2790+ { 0x0400, 0x45, 0x00, 0x00, 0x00 }, /* Ѐ to E */
2791+ { 0x0401, 0x45, 0x00, 0x00, 0x00 }, /* Ё to E */
2792+ { 0x0402, 0x44, 0x00, 0x00, 0x00 }, /* Ђ to D */
2793+ { 0x0403, 0x47, 0x00, 0x00, 0x00 }, /* Ѓ to G */
2794+ { 0x0404, 0x45, 0x00, 0x00, 0x00 }, /* Є to E */
2795+ { 0x0405, 0x5a, 0x00, 0x00, 0x00 }, /* Ѕ to Z */
2796+ { 0x0406, 0x49, 0x00, 0x00, 0x00 }, /* І to I */
2797+ { 0x0407, 0x49, 0x00, 0x00, 0x00 }, /* Ї to I */
2798+ { 0x0408, 0x4a, 0x00, 0x00, 0x00 }, /* Ј to J */
2799+ { 0x0409, 0x49, 0x00, 0x00, 0x00 }, /* Љ to I */
2800+ { 0x040A, 0x4e, 0x00, 0x00, 0x00 }, /* Њ to N */
2801+ { 0x040B, 0x44, 0x00, 0x00, 0x00 }, /* Ћ to D */
2802+ { 0x040C, 0x4b, 0x00, 0x00, 0x00 }, /* Ќ to K */
2803+ { 0x040D, 0x49, 0x00, 0x00, 0x00 }, /* Ѝ to I */
2804+ { 0x040E, 0x55, 0x00, 0x00, 0x00 }, /* Ў to U */
2805+ { 0x040F, 0x44, 0x00, 0x00, 0x00 }, /* Џ to D */
2806+ { 0x0410, 0x41, 0x00, 0x00, 0x00 }, /* А to A */
2807+ { 0x0411, 0x42, 0x00, 0x00, 0x00 }, /* Б to B */
2808+ { 0x0412, 0x56, 0x00, 0x00, 0x00 }, /* В to V */
2809+ { 0x0413, 0x47, 0x00, 0x00, 0x00 }, /* Г to G */
2810+ { 0x0414, 0x44, 0x00, 0x00, 0x00 }, /* Д to D */
2811+ { 0x0415, 0x45, 0x00, 0x00, 0x00 }, /* Е to E */
2812+ { 0x0416, 0x5a, 0x68, 0x00, 0x00 }, /* Ж to Zh */
2813+ { 0x0417, 0x5a, 0x00, 0x00, 0x00 }, /* З to Z */
2814+ { 0x0418, 0x49, 0x00, 0x00, 0x00 }, /* И to I */
2815+ { 0x0419, 0x49, 0x00, 0x00, 0x00 }, /* Й to I */
2816+ { 0x041A, 0x4b, 0x00, 0x00, 0x00 }, /* К to K */
2817+ { 0x041B, 0x4c, 0x00, 0x00, 0x00 }, /* Л to L */
2818+ { 0x041C, 0x4d, 0x00, 0x00, 0x00 }, /* М to M */
2819+ { 0x041D, 0x4e, 0x00, 0x00, 0x00 }, /* Н to N */
2820+ { 0x041E, 0x4f, 0x00, 0x00, 0x00 }, /* О to O */
2821+ { 0x041F, 0x50, 0x00, 0x00, 0x00 }, /* П to P */
2822+ { 0x0420, 0x52, 0x00, 0x00, 0x00 }, /* Р to R */
2823+ { 0x0421, 0x53, 0x00, 0x00, 0x00 }, /* С to S */
2824+ { 0x0422, 0x54, 0x00, 0x00, 0x00 }, /* Т to T */
2825+ { 0x0423, 0x55, 0x00, 0x00, 0x00 }, /* У to U */
2826+ { 0x0424, 0x46, 0x00, 0x00, 0x00 }, /* Ф to F */
2827+ { 0x0425, 0x4b, 0x68, 0x00, 0x00 }, /* Х to Kh */
2828+ { 0x0426, 0x54, 0x63, 0x00, 0x00 }, /* Ц to Tc */
2829+ { 0x0427, 0x43, 0x68, 0x00, 0x00 }, /* Ч to Ch */
2830+ { 0x0428, 0x53, 0x68, 0x00, 0x00 }, /* Ш to Sh */
2831+ { 0x0429, 0x53, 0x68, 0x63, 0x68 }, /* Щ to Shch */
2832+ { 0x042A, 0x61, 0x00, 0x00, 0x00 }, /* to A */
2833+ { 0x042B, 0x59, 0x00, 0x00, 0x00 }, /* Ы to Y */
2834+ { 0x042C, 0x59, 0x00, 0x00, 0x00 }, /* to Y */
2835+ { 0x042D, 0x45, 0x00, 0x00, 0x00 }, /* Э to E */
2836+ { 0x042E, 0x49, 0x75, 0x00, 0x00 }, /* Ю to Iu */
2837+ { 0x042F, 0x49, 0x61, 0x00, 0x00 }, /* Я to Ia */
2838+ { 0x0430, 0x61, 0x00, 0x00, 0x00 }, /* а to a */
2839+ { 0x0431, 0x62, 0x00, 0x00, 0x00 }, /* б to b */
2840+ { 0x0432, 0x76, 0x00, 0x00, 0x00 }, /* в to v */
2841+ { 0x0433, 0x67, 0x00, 0x00, 0x00 }, /* г to g */
2842+ { 0x0434, 0x64, 0x00, 0x00, 0x00 }, /* д to d */
2843+ { 0x0435, 0x65, 0x00, 0x00, 0x00 }, /* е to e */
2844+ { 0x0436, 0x7a, 0x68, 0x00, 0x00 }, /* ж to zh */
2845+ { 0x0437, 0x7a, 0x00, 0x00, 0x00 }, /* з to z */
2846+ { 0x0438, 0x69, 0x00, 0x00, 0x00 }, /* и to i */
2847+ { 0x0439, 0x69, 0x00, 0x00, 0x00 }, /* й to i */
2848+ { 0x043A, 0x6b, 0x00, 0x00, 0x00 }, /* к to k */
2849+ { 0x043B, 0x6c, 0x00, 0x00, 0x00 }, /* л to l */
2850+ { 0x043C, 0x6d, 0x00, 0x00, 0x00 }, /* м to m */
2851+ { 0x043D, 0x6e, 0x00, 0x00, 0x00 }, /* н to n */
2852+ { 0x043E, 0x6f, 0x00, 0x00, 0x00 }, /* о to o */
2853+ { 0x043F, 0x70, 0x00, 0x00, 0x00 }, /* п to p */
2854+ { 0x0440, 0x72, 0x00, 0x00, 0x00 }, /* р to r */
2855+ { 0x0441, 0x73, 0x00, 0x00, 0x00 }, /* с to s */
2856+ { 0x0442, 0x74, 0x00, 0x00, 0x00 }, /* т to t */
2857+ { 0x0443, 0x75, 0x00, 0x00, 0x00 }, /* у to u */
2858+ { 0x0444, 0x66, 0x00, 0x00, 0x00 }, /* ф to f */
2859+ { 0x0445, 0x6b, 0x68, 0x00, 0x00 }, /* х to kh */
2860+ { 0x0446, 0x74, 0x63, 0x00, 0x00 }, /* ц to tc */
2861+ { 0x0447, 0x63, 0x68, 0x00, 0x00 }, /* ч to ch */
2862+ { 0x0448, 0x73, 0x68, 0x00, 0x00 }, /* ш to sh */
2863+ { 0x0449, 0x73, 0x68, 0x63, 0x68 }, /* щ to shch */
2864+ { 0x044A, 0x61, 0x00, 0x00, 0x00 }, /* to a */
2865+ { 0x044B, 0x79, 0x00, 0x00, 0x00 }, /* ы to y */
2866+ { 0x044C, 0x79, 0x00, 0x00, 0x00 }, /* to y */
2867+ { 0x044D, 0x65, 0x00, 0x00, 0x00 }, /* э to e */
2868+ { 0x044E, 0x69, 0x75, 0x00, 0x00 }, /* ю to iu */
2869+ { 0x044F, 0x69, 0x61, 0x00, 0x00 }, /* я to ia */
2870+ { 0x0450, 0x65, 0x00, 0x00, 0x00 }, /* ѐ to e */
2871+ { 0x0451, 0x65, 0x00, 0x00, 0x00 }, /* ё to e */
2872+ { 0x0452, 0x64, 0x00, 0x00, 0x00 }, /* ђ to d */
2873+ { 0x0453, 0x67, 0x00, 0x00, 0x00 }, /* ѓ to g */
2874+ { 0x0454, 0x65, 0x00, 0x00, 0x00 }, /* є to e */
2875+ { 0x0455, 0x7a, 0x00, 0x00, 0x00 }, /* ѕ to z */
2876+ { 0x0456, 0x69, 0x00, 0x00, 0x00 }, /* і to i */
2877+ { 0x0457, 0x69, 0x00, 0x00, 0x00 }, /* ї to i */
2878+ { 0x0458, 0x6a, 0x00, 0x00, 0x00 }, /* ј to j */
2879+ { 0x0459, 0x69, 0x00, 0x00, 0x00 }, /* љ to i */
2880+ { 0x045A, 0x6e, 0x00, 0x00, 0x00 }, /* њ to n */
2881+ { 0x045B, 0x64, 0x00, 0x00, 0x00 }, /* ћ to d */
2882+ { 0x045C, 0x6b, 0x00, 0x00, 0x00 }, /* ќ to k */
2883+ { 0x045D, 0x69, 0x00, 0x00, 0x00 }, /* ѝ to i */
2884+ { 0x045E, 0x75, 0x00, 0x00, 0x00 }, /* ў to u */
2885+ { 0x045F, 0x64, 0x00, 0x00, 0x00 }, /* џ to d */
2886+ { 0x1E02, 0x42, 0x00, 0x00, 0x00 }, /* Ḃ to B */
2887+ { 0x1E03, 0x62, 0x00, 0x00, 0x00 }, /* ḃ to b */
2888+ { 0x1E0A, 0x44, 0x00, 0x00, 0x00 }, /* Ḋ to D */
2889+ { 0x1E0B, 0x64, 0x00, 0x00, 0x00 }, /* ḋ to d */
2890+ { 0x1E1E, 0x46, 0x00, 0x00, 0x00 }, /* Ḟ to F */
2891+ { 0x1E1F, 0x66, 0x00, 0x00, 0x00 }, /* ḟ to f */
2892+ { 0x1E40, 0x4D, 0x00, 0x00, 0x00 }, /* Ṁ to M */
2893+ { 0x1E41, 0x6D, 0x00, 0x00, 0x00 }, /* ṁ to m */
2894+ { 0x1E56, 0x50, 0x00, 0x00, 0x00 }, /* Ṗ to P */
2895+ { 0x1E57, 0x70, 0x00, 0x00, 0x00 }, /* ṗ to p */
2896+ { 0x1E60, 0x53, 0x00, 0x00, 0x00 }, /* Ṡ to S */
2897+ { 0x1E61, 0x73, 0x00, 0x00, 0x00 }, /* ṡ to s */
2898+ { 0x1E6A, 0x54, 0x00, 0x00, 0x00 }, /* Ṫ to T */
2899+ { 0x1E6B, 0x74, 0x00, 0x00, 0x00 }, /* ṫ to t */
2900+ { 0x1E80, 0x57, 0x00, 0x00, 0x00 }, /* Ẁ to W */
2901+ { 0x1E81, 0x77, 0x00, 0x00, 0x00 }, /* ẁ to w */
2902+ { 0x1E82, 0x57, 0x00, 0x00, 0x00 }, /* Ẃ to W */
2903+ { 0x1E83, 0x77, 0x00, 0x00, 0x00 }, /* ẃ to w */
2904+ { 0x1E84, 0x57, 0x00, 0x00, 0x00 }, /* Ẅ to W */
2905+ { 0x1E85, 0x77, 0x00, 0x00, 0x00 }, /* ẅ to w */
2906+ { 0x1EF2, 0x59, 0x00, 0x00, 0x00 }, /* Ỳ to Y */
2907+ { 0x1EF3, 0x79, 0x00, 0x00, 0x00 }, /* ỳ to y */
2908+ { 0xFB00, 0x66, 0x66, 0x00, 0x00 }, /* ff to ff */
2909+ { 0xFB01, 0x66, 0x69, 0x00, 0x00 }, /* fi to fi */
2910+ { 0xFB02, 0x66, 0x6C, 0x00, 0x00 }, /* fl to fl */
2911+ { 0xFB05, 0x73, 0x74, 0x00, 0x00 }, /* ſt to st */
2912+ { 0xFB06, 0x73, 0x74, 0x00, 0x00 }, /* st to st */
2913 };
2914
2915-static const Transliteration *spellfixFindTranslit(int c, int *pxTop){
2916- *pxTop = (sizeof(translit)/sizeof(translit[0])) - 1;
2917- return translit;
2918+static const Transliteration *spellfixFindTranslit(int c, int *pxTop)
2919+{
2920+ *pxTop = (sizeof(translit)/sizeof(translit[0])) - 1;
2921+ return translit;
2922 }
2923
2924 /*
2925@@ -1710,56 +1749,57 @@ **
2926 ** Space to hold the returned string comes from sqlite3_malloc() and
2927 ** should be freed by the caller.
2928 */
2929-static unsigned char *transliterate(const unsigned char *zIn, int nIn){
2930+static unsigned char *transliterate(const unsigned char *zIn, int nIn)
2931+{
2932 #ifdef SQLITE_SPELLFIX_5BYTE_MAPPINGS
2933- unsigned char *zOut = sqlite3_malloc64( nIn*5 + 1 );
2934+ unsigned char *zOut = sqlite3_malloc64( nIn*5 + 1 );
2935 #else
2936- unsigned char *zOut = sqlite3_malloc64( nIn*4 + 1 );
2937+ unsigned char *zOut = sqlite3_malloc64( nIn*4 + 1 );
2938 #endif
2939- int c, sz, nOut;
2940- if( zOut==0 ) return 0;
2941- nOut = 0;
2942- while( nIn>0 ){
2943- c = utf8Read(zIn, nIn, &sz);
2944- zIn += sz;
2945- nIn -= sz;
2946- if( c<=127 ){
2947- zOut[nOut++] = (unsigned char)c;
2948- }else{
2949- int xTop, xBtm, x;
2950- const Transliteration *tbl = spellfixFindTranslit(c, &xTop);
2951- xBtm = 0;
2952- while( xTop>=xBtm ){
2953- x = (xTop + xBtm)/2;
2954- if( tbl[x].cFrom==c ){
2955- zOut[nOut++] = tbl[x].cTo0;
2956- if( tbl[x].cTo1 ){
2957- zOut[nOut++] = tbl[x].cTo1;
2958- if( tbl[x].cTo2 ){
2959- zOut[nOut++] = tbl[x].cTo2;
2960- if( tbl[x].cTo3 ){
2961- zOut[nOut++] = tbl[x].cTo3;
2962+ int c, sz, nOut;
2963+ if( zOut==0 ) return 0;
2964+ nOut = 0;
2965+ while( nIn>0 ) {
2966+ c = utf8Read(zIn, nIn, &sz);
2967+ zIn += sz;
2968+ nIn -= sz;
2969+ if( c<=127 ) {
2970+ zOut[nOut++] = (unsigned char)c;
2971+ } else {
2972+ int xTop, xBtm, x;
2973+ const Transliteration *tbl = spellfixFindTranslit(c, &xTop);
2974+ xBtm = 0;
2975+ while( xTop>=xBtm ) {
2976+ x = (xTop + xBtm)/2;
2977+ if( tbl[x].cFrom==c ) {
2978+ zOut[nOut++] = tbl[x].cTo0;
2979+ if( tbl[x].cTo1 ) {
2980+ zOut[nOut++] = tbl[x].cTo1;
2981+ if( tbl[x].cTo2 ) {
2982+ zOut[nOut++] = tbl[x].cTo2;
2983+ if( tbl[x].cTo3 ) {
2984+ zOut[nOut++] = tbl[x].cTo3;
2985 #ifdef SQLITE_SPELLFIX_5BYTE_MAPPINGS
2986- if( tbl[x].cTo4 ){
2987- zOut[nOut++] = tbl[x].cTo4;
2988+ if( tbl[x].cTo4 ) {
2989+ zOut[nOut++] = tbl[x].cTo4;
2990+ }
2991+#endif /* SQLITE_SPELLFIX_5BYTE_MAPPINGS */
2992+ }
2993+ }
2994+ }
2995+ c = 0;
2996+ break;
2997+ } else if( tbl[x].cFrom>c ) {
2998+ xTop = x-1;
2999+ } else {
3000+ xBtm = x+1;
3001 }
3002-#endif /* SQLITE_SPELLFIX_5BYTE_MAPPINGS */
3003- }
3004 }
3005- }
3006- c = 0;
3007- break;
3008- }else if( tbl[x].cFrom>c ){
3009- xTop = x-1;
3010- }else{
3011- xBtm = x+1;
3012+ if( c ) zOut[nOut++] = '?';
3013 }
3014- }
3015- if( c ) zOut[nOut++] = '?';
3016 }
3017- }
3018- zOut[nOut] = 0;
3019- return zOut;
3020+ zOut[nOut] = 0;
3021+ return zOut;
3022 }
3023
3024 /*
3025@@ -1768,65 +1808,66 @@ ** string that transliterates to an ASCII string nTrans bytes or longer.
3026 ** Or, if the transliteration of the input string is less than nTrans
3027 ** bytes in size, return the number of characters in the input string.
3028 */
3029-static int translen_to_charlen(const char *zIn, int nIn, int nTrans){
3030- int i, c, sz, nOut;
3031- int nChar;
3032+static int translen_to_charlen(const char *zIn, int nIn, int nTrans)
3033+{
3034+ int i, c, sz, nOut;
3035+ int nChar;
3036
3037- i = nOut = 0;
3038- for(nChar=0; i<nIn && nOut<nTrans; nChar++){
3039- c = utf8Read((const unsigned char *)&zIn[i], nIn-i, &sz);
3040- i += sz;
3041+ i = nOut = 0;
3042+ for(nChar=0; i<nIn && nOut<nTrans; nChar++) {
3043+ c = utf8Read((const unsigned char *)&zIn[i], nIn-i, &sz);
3044+ i += sz;
3045
3046- nOut++;
3047- if( c>=128 ){
3048- int xTop, xBtm, x;
3049- const Transliteration *tbl = spellfixFindTranslit(c, &xTop);
3050- xBtm = 0;
3051- while( xTop>=xBtm ){
3052- x = (xTop + xBtm)/2;
3053- if( tbl[x].cFrom==c ){
3054- if( tbl[x].cTo1 ){
3055- nOut++;
3056- if( tbl[x].cTo2 ){
3057- nOut++;
3058- if( tbl[x].cTo3 ){
3059- nOut++;
3060- }
3061+ nOut++;
3062+ if( c>=128 ) {
3063+ int xTop, xBtm, x;
3064+ const Transliteration *tbl = spellfixFindTranslit(c, &xTop);
3065+ xBtm = 0;
3066+ while( xTop>=xBtm ) {
3067+ x = (xTop + xBtm)/2;
3068+ if( tbl[x].cFrom==c ) {
3069+ if( tbl[x].cTo1 ) {
3070+ nOut++;
3071+ if( tbl[x].cTo2 ) {
3072+ nOut++;
3073+ if( tbl[x].cTo3 ) {
3074+ nOut++;
3075+ }
3076+ }
3077+ }
3078+ break;
3079+ } else if( tbl[x].cFrom>c ) {
3080+ xTop = x-1;
3081+ } else {
3082+ xBtm = x+1;
3083+ }
3084 }
3085- }
3086- break;
3087- }else if( tbl[x].cFrom>c ){
3088- xTop = x-1;
3089- }else{
3090- xBtm = x+1;
3091 }
3092- }
3093 }
3094- }
3095
3096- return nChar;
3097+ return nChar;
3098 }
3099-
3100
3101 /*
3102 ** spellfix1_translit(X)
3103 **
3104-** Convert a string that contains non-ASCII Roman characters into
3105+** Convert a string that contains non-ASCII Roman characters into
3106 ** pure ASCII.
3107 */
3108 static void transliterateSqlFunc(
3109- sqlite3_context *context,
3110- int argc,
3111- sqlite3_value **argv
3112-){
3113- const unsigned char *zIn = sqlite3_value_text(argv[0]);
3114- int nIn = sqlite3_value_bytes(argv[0]);
3115- unsigned char *zOut = transliterate(zIn, nIn);
3116- if( zOut==0 ){
3117- sqlite3_result_error_nomem(context);
3118- }else{
3119- sqlite3_result_text(context, (char*)zOut, -1, sqlite3_free);
3120- }
3121+ sqlite3_context *context,
3122+ int argc,
3123+ sqlite3_value **argv
3124+)
3125+{
3126+ const unsigned char *zIn = sqlite3_value_text(argv[0]);
3127+ int nIn = sqlite3_value_bytes(argv[0]);
3128+ unsigned char *zOut = transliterate(zIn, nIn);
3129+ if( zOut==0 ) {
3130+ sqlite3_result_error_nomem(context);
3131+ } else {
3132+ sqlite3_result_text(context, (char*)zOut, -1, sqlite3_free);
3133+ }
3134 }
3135
3136 /*
3137@@ -1846,53 +1887,68 @@ ** two or more of the above scripts or 999 if X contains no characters
3138 ** from any of the above scripts.
3139 */
3140 static void scriptCodeSqlFunc(
3141- sqlite3_context *context,
3142- int argc,
3143- sqlite3_value **argv
3144-){
3145- const unsigned char *zIn = sqlite3_value_text(argv[0]);
3146- int nIn = sqlite3_value_bytes(argv[0]);
3147- int c, sz;
3148- int scriptMask = 0;
3149- int res;
3150- int seenDigit = 0;
3151+ sqlite3_context *context,
3152+ int argc,
3153+ sqlite3_value **argv
3154+)
3155+{
3156+ const unsigned char *zIn = sqlite3_value_text(argv[0]);
3157+ int nIn = sqlite3_value_bytes(argv[0]);
3158+ int c, sz;
3159+ int scriptMask = 0;
3160+ int res;
3161+ int seenDigit = 0;
3162 # define SCRIPT_LATIN 0x0001
3163 # define SCRIPT_CYRILLIC 0x0002
3164 # define SCRIPT_GREEK 0x0004
3165 # define SCRIPT_HEBREW 0x0008
3166 # define SCRIPT_ARABIC 0x0010
3167
3168- while( nIn>0 ){
3169- c = utf8Read(zIn, nIn, &sz);
3170- zIn += sz;
3171- nIn -= sz;
3172- if( c<0x02af ){
3173- if( c>=0x80 || midClass[c&0x7f]<CCLASS_DIGIT ){
3174- scriptMask |= SCRIPT_LATIN;
3175- }else if( c>='0' && c<='9' ){
3176- seenDigit = 1;
3177- }
3178- }else if( c>=0x0400 && c<=0x04ff ){
3179- scriptMask |= SCRIPT_CYRILLIC;
3180- }else if( c>=0x0386 && c<=0x03ce ){
3181- scriptMask |= SCRIPT_GREEK;
3182- }else if( c>=0x0590 && c<=0x05ff ){
3183- scriptMask |= SCRIPT_HEBREW;
3184- }else if( c>=0x0600 && c<=0x06ff ){
3185- scriptMask |= SCRIPT_ARABIC;
3186+ while( nIn>0 ) {
3187+ c = utf8Read(zIn, nIn, &sz);
3188+ zIn += sz;
3189+ nIn -= sz;
3190+ if( c<0x02af ) {
3191+ if( c>=0x80 || midClass[c&0x7f]<CCLASS_DIGIT ) {
3192+ scriptMask |= SCRIPT_LATIN;
3193+ } else if( c>='0' && c<='9' ) {
3194+ seenDigit = 1;
3195+ }
3196+ } else if( c>=0x0400 && c<=0x04ff ) {
3197+ scriptMask |= SCRIPT_CYRILLIC;
3198+ } else if( c>=0x0386 && c<=0x03ce ) {
3199+ scriptMask |= SCRIPT_GREEK;
3200+ } else if( c>=0x0590 && c<=0x05ff ) {
3201+ scriptMask |= SCRIPT_HEBREW;
3202+ } else if( c>=0x0600 && c<=0x06ff ) {
3203+ scriptMask |= SCRIPT_ARABIC;
3204+ }
3205 }
3206- }
3207- if( scriptMask==0 && seenDigit ) scriptMask = SCRIPT_LATIN;
3208- switch( scriptMask ){
3209- case 0: res = 999; break;
3210- case SCRIPT_LATIN: res = 215; break;
3211- case SCRIPT_CYRILLIC: res = 220; break;
3212- case SCRIPT_GREEK: res = 200; break;
3213- case SCRIPT_HEBREW: res = 125; break;
3214- case SCRIPT_ARABIC: res = 160; break;
3215- default: res = 998; break;
3216- }
3217- sqlite3_result_int(context, res);
3218+ if( scriptMask==0 && seenDigit ) scriptMask = SCRIPT_LATIN;
3219+ switch( scriptMask ) {
3220+ case 0:
3221+ res = 999;
3222+ break;
3223+ case SCRIPT_LATIN:
3224+ res = 215;
3225+ break;
3226+ case SCRIPT_CYRILLIC:
3227+ res = 220;
3228+ break;
3229+ case SCRIPT_GREEK:
3230+ res = 200;
3231+ break;
3232+ case SCRIPT_HEBREW:
3233+ res = 125;
3234+ break;
3235+ case SCRIPT_ARABIC:
3236+ res = 160;
3237+ break;
3238+ default:
3239+ res = 998;
3240+ break;
3241+ }
3242+ sqlite3_result_int(context, res);
3243 }
3244
3245 /* End transliterate
3246@@ -1912,37 +1968,37 @@ typedef struct spellfix1_cursor spellfix1_cursor;
3247
3248 /* Fuzzy-search virtual table object */
3249 struct spellfix1_vtab {
3250- sqlite3_vtab base; /* Base class - must be first */
3251- sqlite3 *db; /* Database connection */
3252- char *zDbName; /* Name of database holding this table */
3253- char *zTableName; /* Name of the virtual table */
3254- char *zCostTable; /* Table holding edit-distance cost numbers */
3255- EditDist3Config *pConfig3; /* Parsed edit distance costs */
3256+ sqlite3_vtab base; /* Base class - must be first */
3257+ sqlite3 *db; /* Database connection */
3258+ char *zDbName; /* Name of database holding this table */
3259+ char *zTableName; /* Name of the virtual table */
3260+ char *zCostTable; /* Table holding edit-distance cost numbers */
3261+ EditDist3Config *pConfig3; /* Parsed edit distance costs */
3262 };
3263
3264 /* Fuzzy-search cursor object */
3265 struct spellfix1_cursor {
3266- sqlite3_vtab_cursor base; /* Base class - must be first */
3267- spellfix1_vtab *pVTab; /* The table to which this cursor belongs */
3268- char *zPattern; /* rhs of MATCH clause */
3269- int idxNum; /* idxNum value passed to xFilter() */
3270- int nRow; /* Number of rows of content */
3271- int nAlloc; /* Number of allocated rows */
3272- int iRow; /* Current row of content */
3273- int iLang; /* Value of the langid= constraint */
3274- int iTop; /* Value of the top= constraint */
3275- int iScope; /* Value of the scope= constraint */
3276- int nSearch; /* Number of vocabulary items checked */
3277- sqlite3_stmt *pFullScan; /* Shadow query for a full table scan */
3278- struct spellfix1_row { /* For each row of content */
3279- sqlite3_int64 iRowid; /* Rowid for this row */
3280- char *zWord; /* Text for this row */
3281- int iRank; /* Rank for this row */
3282- int iDistance; /* Distance from pattern for this row */
3283- int iScore; /* Score for sorting */
3284- int iMatchlen; /* Value of matchlen column (or -1) */
3285- char zHash[SPELLFIX_MX_HASH]; /* the phonehash used for this match */
3286- } *a;
3287+ sqlite3_vtab_cursor base; /* Base class - must be first */
3288+ spellfix1_vtab *pVTab; /* The table to which this cursor belongs */
3289+ char *zPattern; /* rhs of MATCH clause */
3290+ int idxNum; /* idxNum value passed to xFilter() */
3291+ int nRow; /* Number of rows of content */
3292+ int nAlloc; /* Number of allocated rows */
3293+ int iRow; /* Current row of content */
3294+ int iLang; /* Value of the langid= constraint */
3295+ int iTop; /* Value of the top= constraint */
3296+ int iScope; /* Value of the scope= constraint */
3297+ int nSearch; /* Number of vocabulary items checked */
3298+ sqlite3_stmt *pFullScan; /* Shadow query for a full table scan */
3299+ struct spellfix1_row { /* For each row of content */
3300+ sqlite3_int64 iRowid; /* Rowid for this row */
3301+ char *zWord; /* Text for this row */
3302+ int iRank; /* Rank for this row */
3303+ int iDistance; /* Distance from pattern for this row */
3304+ int iScore; /* Score for sorting */
3305+ int iMatchlen; /* Value of matchlen column (or -1) */
3306+ char zHash[SPELLFIX_MX_HASH]; /* the phonehash used for this match */
3307+ } *a;
3308 };
3309
3310 /*
3311@@ -1953,84 +2009,90 @@ **
3312 ** If *pRc is initially non-zero then this routine is a no-op.
3313 */
3314 static void spellfix1DbExec(
3315- int *pRc, /* Success code */
3316- sqlite3 *db, /* Database in which to run SQL */
3317- const char *zFormat, /* Format string for SQL */
3318- ... /* Arguments to the format string */
3319-){
3320- va_list ap;
3321- char *zSql;
3322- if( *pRc ) return;
3323- va_start(ap, zFormat);
3324- zSql = sqlite3_vmprintf(zFormat, ap);
3325- va_end(ap);
3326- if( zSql==0 ){
3327- *pRc = SQLITE_NOMEM;
3328- }else{
3329- *pRc = sqlite3_exec(db, zSql, 0, 0, 0);
3330- sqlite3_free(zSql);
3331- }
3332+ int *pRc, /* Success code */
3333+ sqlite3 *db, /* Database in which to run SQL */
3334+ const char *zFormat, /* Format string for SQL */
3335+ ... /* Arguments to the format string */
3336+)
3337+{
3338+ va_list ap;
3339+ char *zSql;
3340+ if( *pRc ) return;
3341+ va_start(ap, zFormat);
3342+ zSql = sqlite3_vmprintf(zFormat, ap);
3343+ va_end(ap);
3344+ if( zSql==0 ) {
3345+ *pRc = SQLITE_NOMEM;
3346+ } else {
3347+ *pRc = sqlite3_exec(db, zSql, 0, 0, 0);
3348+ sqlite3_free(zSql);
3349+ }
3350 }
3351
3352 /*
3353 ** xDisconnect/xDestroy method for the fuzzy-search module.
3354 */
3355-static int spellfix1Uninit(int isDestroy, sqlite3_vtab *pVTab){
3356- spellfix1_vtab *p = (spellfix1_vtab*)pVTab;
3357- int rc = SQLITE_OK;
3358- if( isDestroy ){
3359- sqlite3 *db = p->db;
3360- spellfix1DbExec(&rc, db, "DROP TABLE IF EXISTS \"%w\".\"%w_vocab\"",
3361- p->zDbName, p->zTableName);
3362- }
3363- if( rc==SQLITE_OK ){
3364- sqlite3_free(p->zTableName);
3365- editDist3ConfigDelete(p->pConfig3);
3366- sqlite3_free(p->zCostTable);
3367- sqlite3_free(p);
3368- }
3369- return rc;
3370+static int spellfix1Uninit(int isDestroy, sqlite3_vtab *pVTab)
3371+{
3372+ spellfix1_vtab *p = (spellfix1_vtab*)pVTab;
3373+ int rc = SQLITE_OK;
3374+ if( isDestroy ) {
3375+ sqlite3 *db = p->db;
3376+ spellfix1DbExec(&rc, db, "DROP TABLE IF EXISTS \"%w\".\"%w_vocab\"",
3377+ p->zDbName, p->zTableName);
3378+ }
3379+ if( rc==SQLITE_OK ) {
3380+ sqlite3_free(p->zTableName);
3381+ editDist3ConfigDelete(p->pConfig3);
3382+ sqlite3_free(p->zCostTable);
3383+ sqlite3_free(p);
3384+ }
3385+ return rc;
3386 }
3387-static int spellfix1Disconnect(sqlite3_vtab *pVTab){
3388- return spellfix1Uninit(0, pVTab);
3389+static int spellfix1Disconnect(sqlite3_vtab *pVTab)
3390+{
3391+ return spellfix1Uninit(0, pVTab);
3392 }
3393-static int spellfix1Destroy(sqlite3_vtab *pVTab){
3394- return spellfix1Uninit(1, pVTab);
3395+static int spellfix1Destroy(sqlite3_vtab *pVTab)
3396+{
3397+ return spellfix1Uninit(1, pVTab);
3398 }
3399
3400 /*
3401 ** Make a copy of a string. Remove leading and trailing whitespace
3402 ** and dequote it.
3403 */
3404-static char *spellfix1Dequote(const char *zIn){
3405- char *zOut;
3406- int i, j;
3407- char c;
3408- while( isspace((unsigned char)zIn[0]) ) zIn++;
3409- zOut = sqlite3_mprintf("%s", zIn);
3410- if( zOut==0 ) return 0;
3411- i = (int)strlen(zOut);
3412+static char *spellfix1Dequote(const char *zIn)
3413+{
3414+ char *zOut;
3415+ int i, j;
3416+ char c;
3417+ while( isspace((unsigned char)zIn[0]) ) zIn++;
3418+ zOut = sqlite3_mprintf("%s", zIn);
3419+ if( zOut==0 ) return 0;
3420+ i = (int)strlen(zOut);
3421 #if 0 /* The parser will never leave spaces at the end */
3422- while( i>0 && isspace(zOut[i-1]) ){ i--; }
3423+ while( i>0 && isspace(zOut[i-1]) ) {
3424+ i--;
3425+ }
3426 #endif
3427- zOut[i] = 0;
3428- c = zOut[0];
3429- if( c=='\'' || c=='"' ){
3430- for(i=1, j=0; ALWAYS(zOut[i]); i++){
3431- zOut[j++] = zOut[i];
3432- if( zOut[i]==c ){
3433- if( zOut[i+1]==c ){
3434- i++;
3435- }else{
3436- zOut[j-1] = 0;
3437- break;
3438+ zOut[i] = 0;
3439+ c = zOut[0];
3440+ if( c=='\'' || c=='"' ) {
3441+ for(i=1, j=0; ALWAYS(zOut[i]); i++) {
3442+ zOut[j++] = zOut[i];
3443+ if( zOut[i]==c ) {
3444+ if( zOut[i+1]==c ) {
3445+ i++;
3446+ } else {
3447+ zOut[j-1] = 0;
3448+ break;
3449+ }
3450+ }
3451 }
3452- }
3453 }
3454- }
3455- return zOut;
3456+ return zOut;
3457 }
3458-
3459
3460 /*
3461 ** xConnect/xCreate method for the spellfix1 module. Arguments are:
3462@@ -2041,41 +2103,42 @@ ** argv[2] -> table name
3463 ** argv[3].. -> optional arguments (i.e. "edit_cost_table" parameter)
3464 */
3465 static int spellfix1Init(
3466- int isCreate,
3467- sqlite3 *db,
3468- void *pAux,
3469- int argc, const char *const*argv,
3470- sqlite3_vtab **ppVTab,
3471- char **pzErr
3472-){
3473- spellfix1_vtab *pNew = 0;
3474- /* const char *zModule = argv[0]; // not used */
3475- const char *zDbName = argv[1];
3476- const char *zTableName = argv[2];
3477- int nDbName;
3478- int rc = SQLITE_OK;
3479- int i;
3480+ int isCreate,
3481+ sqlite3 *db,
3482+ void *pAux,
3483+ int argc, const char *const*argv,
3484+ sqlite3_vtab **ppVTab,
3485+ char **pzErr
3486+)
3487+{
3488+ spellfix1_vtab *pNew = 0;
3489+ /* const char *zModule = argv[0]; // not used */
3490+ const char *zDbName = argv[1];
3491+ const char *zTableName = argv[2];
3492+ int nDbName;
3493+ int rc = SQLITE_OK;
3494+ int i;
3495
3496- nDbName = (int)strlen(zDbName);
3497- pNew = sqlite3_malloc64( sizeof(*pNew) + nDbName + 1);
3498- if( pNew==0 ){
3499- rc = SQLITE_NOMEM;
3500- }else{
3501- memset(pNew, 0, sizeof(*pNew));
3502- pNew->zDbName = (char*)&pNew[1];
3503- memcpy(pNew->zDbName, zDbName, nDbName+1);
3504- pNew->zTableName = sqlite3_mprintf("%s", zTableName);
3505- pNew->db = db;
3506- if( pNew->zTableName==0 ){
3507- rc = SQLITE_NOMEM;
3508- }else{
3509- sqlite3_vtab_config(db, SQLITE_VTAB_INNOCUOUS);
3510- rc = sqlite3_declare_vtab(db,
3511- "CREATE TABLE x(word,rank,distance,langid, "
3512- "score, matchlen, phonehash HIDDEN, "
3513- "top HIDDEN, scope HIDDEN, srchcnt HIDDEN, "
3514- "soundslike HIDDEN, command HIDDEN)"
3515- );
3516+ nDbName = (int)strlen(zDbName);
3517+ pNew = sqlite3_malloc64( sizeof(*pNew) + nDbName + 1);
3518+ if( pNew==0 ) {
3519+ rc = SQLITE_NOMEM;
3520+ } else {
3521+ memset(pNew, 0, sizeof(*pNew));
3522+ pNew->zDbName = (char*)&pNew[1];
3523+ memcpy(pNew->zDbName, zDbName, nDbName+1);
3524+ pNew->zTableName = sqlite3_mprintf("%s", zTableName);
3525+ pNew->db = db;
3526+ if( pNew->zTableName==0 ) {
3527+ rc = SQLITE_NOMEM;
3528+ } else {
3529+ sqlite3_vtab_config(db, SQLITE_VTAB_INNOCUOUS);
3530+ rc = sqlite3_declare_vtab(db,
3531+ "CREATE TABLE x(word,rank,distance,langid, "
3532+ "score, matchlen, phonehash HIDDEN, "
3533+ "top HIDDEN, scope HIDDEN, srchcnt HIDDEN, "
3534+ "soundslike HIDDEN, command HIDDEN)"
3535+ );
3536 #define SPELLFIX_COL_WORD 0
3537 #define SPELLFIX_COL_RANK 1
3538 #define SPELLFIX_COL_DISTANCE 2
3539@@ -2088,113 +2151,117 @@ #define SPELLFIX_COL_SCOPE 8
3540 #define SPELLFIX_COL_SRCHCNT 9
3541 #define SPELLFIX_COL_SOUNDSLIKE 10
3542 #define SPELLFIX_COL_COMMAND 11
3543+ }
3544+ if( rc==SQLITE_OK && isCreate ) {
3545+ spellfix1DbExec(&rc, db,
3546+ "CREATE TABLE IF NOT EXISTS \"%w\".\"%w_vocab\"(\n"
3547+ " id INTEGER PRIMARY KEY,\n"
3548+ " rank INT,\n"
3549+ " langid INT,\n"
3550+ " word TEXT,\n"
3551+ " k1 TEXT,\n"
3552+ " k2 TEXT\n"
3553+ ");\n",
3554+ zDbName, zTableName
3555+ );
3556+ spellfix1DbExec(&rc, db,
3557+ "CREATE INDEX IF NOT EXISTS \"%w\".\"%w_vocab_index_langid_k2\" "
3558+ "ON \"%w_vocab\"(langid,k2);",
3559+ zDbName, zTableName, zTableName
3560+ );
3561+ }
3562+ for(i=3; rc==SQLITE_OK && i<argc; i++) {
3563+ if( strncmp(argv[i],"edit_cost_table=",16)==0 && pNew->zCostTable==0 ) {
3564+ pNew->zCostTable = spellfix1Dequote(&argv[i][16]);
3565+ if( pNew->zCostTable==0 ) rc = SQLITE_NOMEM;
3566+ continue;
3567+ }
3568+ *pzErr = sqlite3_mprintf("bad argument to spellfix1(): \"%s\"", argv[i]);
3569+ rc = SQLITE_ERROR;
3570+ }
3571 }
3572- if( rc==SQLITE_OK && isCreate ){
3573- spellfix1DbExec(&rc, db,
3574- "CREATE TABLE IF NOT EXISTS \"%w\".\"%w_vocab\"(\n"
3575- " id INTEGER PRIMARY KEY,\n"
3576- " rank INT,\n"
3577- " langid INT,\n"
3578- " word TEXT,\n"
3579- " k1 TEXT,\n"
3580- " k2 TEXT\n"
3581- ");\n",
3582- zDbName, zTableName
3583- );
3584- spellfix1DbExec(&rc, db,
3585- "CREATE INDEX IF NOT EXISTS \"%w\".\"%w_vocab_index_langid_k2\" "
3586- "ON \"%w_vocab\"(langid,k2);",
3587- zDbName, zTableName, zTableName
3588- );
3589+
3590+ if( rc && pNew ) {
3591+ *ppVTab = 0;
3592+ spellfix1Uninit(0, &pNew->base);
3593+ } else {
3594+ *ppVTab = (sqlite3_vtab *)pNew;
3595 }
3596- for(i=3; rc==SQLITE_OK && i<argc; i++){
3597- if( strncmp(argv[i],"edit_cost_table=",16)==0 && pNew->zCostTable==0 ){
3598- pNew->zCostTable = spellfix1Dequote(&argv[i][16]);
3599- if( pNew->zCostTable==0 ) rc = SQLITE_NOMEM;
3600- continue;
3601- }
3602- *pzErr = sqlite3_mprintf("bad argument to spellfix1(): \"%s\"", argv[i]);
3603- rc = SQLITE_ERROR;
3604- }
3605- }
3606-
3607- if( rc && pNew ){
3608- *ppVTab = 0;
3609- spellfix1Uninit(0, &pNew->base);
3610- }else{
3611- *ppVTab = (sqlite3_vtab *)pNew;
3612- }
3613- return rc;
3614+ return rc;
3615 }
3616
3617 /*
3618 ** The xConnect and xCreate methods
3619 */
3620 static int spellfix1Connect(
3621- sqlite3 *db,
3622- void *pAux,
3623- int argc, const char *const*argv,
3624- sqlite3_vtab **ppVTab,
3625- char **pzErr
3626-){
3627- return spellfix1Init(0, db, pAux, argc, argv, ppVTab, pzErr);
3628+ sqlite3 *db,
3629+ void *pAux,
3630+ int argc, const char *const*argv,
3631+ sqlite3_vtab **ppVTab,
3632+ char **pzErr
3633+)
3634+{
3635+ return spellfix1Init(0, db, pAux, argc, argv, ppVTab, pzErr);
3636 }
3637 static int spellfix1Create(
3638- sqlite3 *db,
3639- void *pAux,
3640- int argc, const char *const*argv,
3641- sqlite3_vtab **ppVTab,
3642- char **pzErr
3643-){
3644- return spellfix1Init(1, db, pAux, argc, argv, ppVTab, pzErr);
3645+ sqlite3 *db,
3646+ void *pAux,
3647+ int argc, const char *const*argv,
3648+ sqlite3_vtab **ppVTab,
3649+ char **pzErr
3650+)
3651+{
3652+ return spellfix1Init(1, db, pAux, argc, argv, ppVTab, pzErr);
3653 }
3654
3655 /*
3656 ** Clear all of the content from a cursor.
3657 */
3658-static void spellfix1ResetCursor(spellfix1_cursor *pCur){
3659- int i;
3660- for(i=0; i<pCur->nRow; i++){
3661- sqlite3_free(pCur->a[i].zWord);
3662- }
3663- pCur->nRow = 0;
3664- pCur->iRow = 0;
3665- pCur->nSearch = 0;
3666- if( pCur->pFullScan ){
3667- sqlite3_finalize(pCur->pFullScan);
3668- pCur->pFullScan = 0;
3669- }
3670+static void spellfix1ResetCursor(spellfix1_cursor *pCur)
3671+{
3672+ int i;
3673+ for(i=0; i<pCur->nRow; i++) {
3674+ sqlite3_free(pCur->a[i].zWord);
3675+ }
3676+ pCur->nRow = 0;
3677+ pCur->iRow = 0;
3678+ pCur->nSearch = 0;
3679+ if( pCur->pFullScan ) {
3680+ sqlite3_finalize(pCur->pFullScan);
3681+ pCur->pFullScan = 0;
3682+ }
3683 }
3684
3685 /*
3686 ** Resize the cursor to hold up to N rows of content
3687 */
3688-static void spellfix1ResizeCursor(spellfix1_cursor *pCur, int N){
3689- struct spellfix1_row *aNew;
3690- assert( N>=pCur->nRow );
3691- aNew = sqlite3_realloc64(pCur->a, sizeof(pCur->a[0])*N);
3692- if( aNew==0 && N>0 ){
3693- spellfix1ResetCursor(pCur);
3694- sqlite3_free(pCur->a);
3695- pCur->nAlloc = 0;
3696- pCur->a = 0;
3697- }else{
3698- pCur->nAlloc = N;
3699- pCur->a = aNew;
3700- }
3701+static void spellfix1ResizeCursor(spellfix1_cursor *pCur, int N)
3702+{
3703+ struct spellfix1_row *aNew;
3704+ assert( N>=pCur->nRow );
3705+ aNew = sqlite3_realloc64(pCur->a, sizeof(pCur->a[0])*N);
3706+ if( aNew==0 && N>0 ) {
3707+ spellfix1ResetCursor(pCur);
3708+ sqlite3_free(pCur->a);
3709+ pCur->nAlloc = 0;
3710+ pCur->a = 0;
3711+ } else {
3712+ pCur->nAlloc = N;
3713+ pCur->a = aNew;
3714+ }
3715 }
3716-
3717
3718 /*
3719 ** Close a fuzzy-search cursor.
3720 */
3721-static int spellfix1Close(sqlite3_vtab_cursor *cur){
3722- spellfix1_cursor *pCur = (spellfix1_cursor *)cur;
3723- spellfix1ResetCursor(pCur);
3724- spellfix1ResizeCursor(pCur, 0);
3725- sqlite3_free(pCur->zPattern);
3726- sqlite3_free(pCur);
3727- return SQLITE_OK;
3728+static int spellfix1Close(sqlite3_vtab_cursor *cur)
3729+{
3730+ spellfix1_cursor *pCur = (spellfix1_cursor *)cur;
3731+ spellfix1ResetCursor(pCur);
3732+ spellfix1ResizeCursor(pCur, 0);
3733+ sqlite3_free(pCur->zPattern);
3734+ sqlite3_free(pCur);
3735+ return SQLITE_OK;
3736 }
3737
3738 #define SPELLFIX_IDXNUM_MATCH 0x01 /* word MATCH $str */
3739@@ -2214,149 +2281,153 @@ **
3740 ** filter.argv[*] values contains $str, $langid, $top, $scope and $rowid
3741 ** if specified and in that order.
3742 */
3743-static int spellfix1BestIndex(sqlite3_vtab *tab, sqlite3_index_info *pIdxInfo){
3744- int iPlan = 0;
3745- int iLangTerm = -1;
3746- int iTopTerm = -1;
3747- int iScopeTerm = -1;
3748- int iDistTerm = -1;
3749- int iRowidTerm = -1;
3750- int i;
3751- const struct sqlite3_index_constraint *pConstraint;
3752- pConstraint = pIdxInfo->aConstraint;
3753- for(i=0; i<pIdxInfo->nConstraint; i++, pConstraint++){
3754- if( pConstraint->usable==0 ) continue;
3755+static int spellfix1BestIndex(sqlite3_vtab *tab, sqlite3_index_info *pIdxInfo)
3756+{
3757+ int iPlan = 0;
3758+ int iLangTerm = -1;
3759+ int iTopTerm = -1;
3760+ int iScopeTerm = -1;
3761+ int iDistTerm = -1;
3762+ int iRowidTerm = -1;
3763+ int i;
3764+ const struct sqlite3_index_constraint *pConstraint;
3765+ pConstraint = pIdxInfo->aConstraint;
3766+ for(i=0; i<pIdxInfo->nConstraint; i++, pConstraint++) {
3767+ if( pConstraint->usable==0 ) continue;
3768
3769- /* Terms of the form: word MATCH $str */
3770- if( (iPlan & SPELLFIX_IDXNUM_MATCH)==0
3771- && pConstraint->iColumn==SPELLFIX_COL_WORD
3772- && pConstraint->op==SQLITE_INDEX_CONSTRAINT_MATCH
3773- ){
3774- iPlan |= SPELLFIX_IDXNUM_MATCH;
3775- pIdxInfo->aConstraintUsage[i].argvIndex = 1;
3776- pIdxInfo->aConstraintUsage[i].omit = 1;
3777- }
3778+ /* Terms of the form: word MATCH $str */
3779+ if( (iPlan & SPELLFIX_IDXNUM_MATCH)==0
3780+ && pConstraint->iColumn==SPELLFIX_COL_WORD
3781+ && pConstraint->op==SQLITE_INDEX_CONSTRAINT_MATCH
3782+ ) {
3783+ iPlan |= SPELLFIX_IDXNUM_MATCH;
3784+ pIdxInfo->aConstraintUsage[i].argvIndex = 1;
3785+ pIdxInfo->aConstraintUsage[i].omit = 1;
3786+ }
3787
3788- /* Terms of the form: langid = $langid */
3789- if( (iPlan & SPELLFIX_IDXNUM_LANGID)==0
3790- && pConstraint->iColumn==SPELLFIX_COL_LANGID
3791- && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ
3792- ){
3793- iPlan |= SPELLFIX_IDXNUM_LANGID;
3794- iLangTerm = i;
3795- }
3796+ /* Terms of the form: langid = $langid */
3797+ if( (iPlan & SPELLFIX_IDXNUM_LANGID)==0
3798+ && pConstraint->iColumn==SPELLFIX_COL_LANGID
3799+ && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ
3800+ ) {
3801+ iPlan |= SPELLFIX_IDXNUM_LANGID;
3802+ iLangTerm = i;
3803+ }
3804
3805- /* Terms of the form: top = $top */
3806- if( (iPlan & SPELLFIX_IDXNUM_TOP)==0
3807- && pConstraint->iColumn==SPELLFIX_COL_TOP
3808- && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ
3809- ){
3810- iPlan |= SPELLFIX_IDXNUM_TOP;
3811- iTopTerm = i;
3812- }
3813+ /* Terms of the form: top = $top */
3814+ if( (iPlan & SPELLFIX_IDXNUM_TOP)==0
3815+ && pConstraint->iColumn==SPELLFIX_COL_TOP
3816+ && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ
3817+ ) {
3818+ iPlan |= SPELLFIX_IDXNUM_TOP;
3819+ iTopTerm = i;
3820+ }
3821
3822- /* Terms of the form: scope = $scope */
3823- if( (iPlan & SPELLFIX_IDXNUM_SCOPE)==0
3824- && pConstraint->iColumn==SPELLFIX_COL_SCOPE
3825- && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ
3826- ){
3827- iPlan |= SPELLFIX_IDXNUM_SCOPE;
3828- iScopeTerm = i;
3829- }
3830+ /* Terms of the form: scope = $scope */
3831+ if( (iPlan & SPELLFIX_IDXNUM_SCOPE)==0
3832+ && pConstraint->iColumn==SPELLFIX_COL_SCOPE
3833+ && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ
3834+ ) {
3835+ iPlan |= SPELLFIX_IDXNUM_SCOPE;
3836+ iScopeTerm = i;
3837+ }
3838
3839- /* Terms of the form: distance < $dist or distance <= $dist */
3840- if( (iPlan & SPELLFIX_IDXNUM_DIST)==0
3841- && pConstraint->iColumn==SPELLFIX_COL_DISTANCE
3842- && (pConstraint->op==SQLITE_INDEX_CONSTRAINT_LT
3843- || pConstraint->op==SQLITE_INDEX_CONSTRAINT_LE)
3844- ){
3845- if( pConstraint->op==SQLITE_INDEX_CONSTRAINT_LT ){
3846- iPlan |= SPELLFIX_IDXNUM_DISTLT;
3847- }else{
3848- iPlan |= SPELLFIX_IDXNUM_DISTLE;
3849- }
3850- iDistTerm = i;
3851- }
3852+ /* Terms of the form: distance < $dist or distance <= $dist */
3853+ if( (iPlan & SPELLFIX_IDXNUM_DIST)==0
3854+ && pConstraint->iColumn==SPELLFIX_COL_DISTANCE
3855+ && (pConstraint->op==SQLITE_INDEX_CONSTRAINT_LT
3856+ || pConstraint->op==SQLITE_INDEX_CONSTRAINT_LE)
3857+ ) {
3858+ if( pConstraint->op==SQLITE_INDEX_CONSTRAINT_LT ) {
3859+ iPlan |= SPELLFIX_IDXNUM_DISTLT;
3860+ } else {
3861+ iPlan |= SPELLFIX_IDXNUM_DISTLE;
3862+ }
3863+ iDistTerm = i;
3864+ }
3865
3866- /* Terms of the form: distance < $dist or distance <= $dist */
3867- if( (iPlan & SPELLFIX_IDXNUM_ROWID)==0
3868- && pConstraint->iColumn<0
3869- && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ
3870- ){
3871- iPlan |= SPELLFIX_IDXNUM_ROWID;
3872- iRowidTerm = i;
3873+ /* Terms of the form: distance < $dist or distance <= $dist */
3874+ if( (iPlan & SPELLFIX_IDXNUM_ROWID)==0
3875+ && pConstraint->iColumn<0
3876+ && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ
3877+ ) {
3878+ iPlan |= SPELLFIX_IDXNUM_ROWID;
3879+ iRowidTerm = i;
3880+ }
3881 }
3882- }
3883- if( iPlan&SPELLFIX_IDXNUM_MATCH ){
3884- int idx = 2;
3885- pIdxInfo->idxNum = iPlan;
3886- if( pIdxInfo->nOrderBy==1
3887- && pIdxInfo->aOrderBy[0].iColumn==SPELLFIX_COL_SCORE
3888- && pIdxInfo->aOrderBy[0].desc==0
3889- ){
3890- pIdxInfo->orderByConsumed = 1; /* Default order by iScore */
3891+ if( iPlan&SPELLFIX_IDXNUM_MATCH ) {
3892+ int idx = 2;
3893+ pIdxInfo->idxNum = iPlan;
3894+ if( pIdxInfo->nOrderBy==1
3895+ && pIdxInfo->aOrderBy[0].iColumn==SPELLFIX_COL_SCORE
3896+ && pIdxInfo->aOrderBy[0].desc==0
3897+ ) {
3898+ pIdxInfo->orderByConsumed = 1; /* Default order by iScore */
3899+ }
3900+ if( iPlan&SPELLFIX_IDXNUM_LANGID ) {
3901+ pIdxInfo->aConstraintUsage[iLangTerm].argvIndex = idx++;
3902+ pIdxInfo->aConstraintUsage[iLangTerm].omit = 1;
3903+ }
3904+ if( iPlan&SPELLFIX_IDXNUM_TOP ) {
3905+ pIdxInfo->aConstraintUsage[iTopTerm].argvIndex = idx++;
3906+ pIdxInfo->aConstraintUsage[iTopTerm].omit = 1;
3907+ }
3908+ if( iPlan&SPELLFIX_IDXNUM_SCOPE ) {
3909+ pIdxInfo->aConstraintUsage[iScopeTerm].argvIndex = idx++;
3910+ pIdxInfo->aConstraintUsage[iScopeTerm].omit = 1;
3911+ }
3912+ if( iPlan&SPELLFIX_IDXNUM_DIST ) {
3913+ pIdxInfo->aConstraintUsage[iDistTerm].argvIndex = idx++;
3914+ pIdxInfo->aConstraintUsage[iDistTerm].omit = 1;
3915+ }
3916+ pIdxInfo->estimatedCost = 1e5;
3917+ } else if( (iPlan & SPELLFIX_IDXNUM_ROWID) ) {
3918+ pIdxInfo->idxNum = SPELLFIX_IDXNUM_ROWID;
3919+ pIdxInfo->aConstraintUsage[iRowidTerm].argvIndex = 1;
3920+ pIdxInfo->aConstraintUsage[iRowidTerm].omit = 1;
3921+ pIdxInfo->estimatedCost = 5;
3922+ } else {
3923+ pIdxInfo->idxNum = 0;
3924+ pIdxInfo->estimatedCost = 1e50;
3925 }
3926- if( iPlan&SPELLFIX_IDXNUM_LANGID ){
3927- pIdxInfo->aConstraintUsage[iLangTerm].argvIndex = idx++;
3928- pIdxInfo->aConstraintUsage[iLangTerm].omit = 1;
3929- }
3930- if( iPlan&SPELLFIX_IDXNUM_TOP ){
3931- pIdxInfo->aConstraintUsage[iTopTerm].argvIndex = idx++;
3932- pIdxInfo->aConstraintUsage[iTopTerm].omit = 1;
3933- }
3934- if( iPlan&SPELLFIX_IDXNUM_SCOPE ){
3935- pIdxInfo->aConstraintUsage[iScopeTerm].argvIndex = idx++;
3936- pIdxInfo->aConstraintUsage[iScopeTerm].omit = 1;
3937- }
3938- if( iPlan&SPELLFIX_IDXNUM_DIST ){
3939- pIdxInfo->aConstraintUsage[iDistTerm].argvIndex = idx++;
3940- pIdxInfo->aConstraintUsage[iDistTerm].omit = 1;
3941- }
3942- pIdxInfo->estimatedCost = 1e5;
3943- }else if( (iPlan & SPELLFIX_IDXNUM_ROWID) ){
3944- pIdxInfo->idxNum = SPELLFIX_IDXNUM_ROWID;
3945- pIdxInfo->aConstraintUsage[iRowidTerm].argvIndex = 1;
3946- pIdxInfo->aConstraintUsage[iRowidTerm].omit = 1;
3947- pIdxInfo->estimatedCost = 5;
3948- }else{
3949- pIdxInfo->idxNum = 0;
3950- pIdxInfo->estimatedCost = 1e50;
3951- }
3952- return SQLITE_OK;
3953+ return SQLITE_OK;
3954 }
3955
3956 /*
3957 ** Open a new fuzzy-search cursor.
3958 */
3959-static int spellfix1Open(sqlite3_vtab *pVTab, sqlite3_vtab_cursor **ppCursor){
3960- spellfix1_vtab *p = (spellfix1_vtab*)pVTab;
3961- spellfix1_cursor *pCur;
3962- pCur = sqlite3_malloc64( sizeof(*pCur) );
3963- if( pCur==0 ) return SQLITE_NOMEM;
3964- memset(pCur, 0, sizeof(*pCur));
3965- pCur->pVTab = p;
3966- *ppCursor = &pCur->base;
3967- return SQLITE_OK;
3968+static int spellfix1Open(sqlite3_vtab *pVTab, sqlite3_vtab_cursor **ppCursor)
3969+{
3970+ spellfix1_vtab *p = (spellfix1_vtab*)pVTab;
3971+ spellfix1_cursor *pCur;
3972+ pCur = sqlite3_malloc64( sizeof(*pCur) );
3973+ if( pCur==0 ) return SQLITE_NOMEM;
3974+ memset(pCur, 0, sizeof(*pCur));
3975+ pCur->pVTab = p;
3976+ *ppCursor = &pCur->base;
3977+ return SQLITE_OK;
3978 }
3979
3980 /*
3981 ** Adjust a distance measurement by the words rank in order to show
3982 ** preference to common words.
3983 */
3984-static int spellfix1Score(int iDistance, int iRank){
3985- int iLog2;
3986- for(iLog2=0; iRank>0; iLog2++, iRank>>=1){}
3987- return iDistance + 32 - iLog2;
3988+static int spellfix1Score(int iDistance, int iRank)
3989+{
3990+ int iLog2;
3991+ for(iLog2=0; iRank>0; iLog2++, iRank>>=1) {}
3992+ return iDistance + 32 - iLog2;
3993 }
3994
3995 /*
3996 ** Compare two spellfix1_row objects for sorting purposes in qsort() such
3997 ** that they sort in order of increasing distance.
3998 */
3999-static int SQLITE_CDECL spellfix1RowCompare(const void *A, const void *B){
4000- const struct spellfix1_row *a = (const struct spellfix1_row*)A;
4001- const struct spellfix1_row *b = (const struct spellfix1_row*)B;
4002- return a->iScore - b->iScore;
4003+static int SQLITE_CDECL spellfix1RowCompare(const void *A, const void *B)
4004+{
4005+ const struct spellfix1_row *a = (const struct spellfix1_row*)A;
4006+ const struct spellfix1_row *b = (const struct spellfix1_row*)B;
4007+ return a->iScore - b->iScore;
4008 }
4009
4010 /*
4011@@ -2364,159 +2435,160 @@ ** A structure used to pass information from spellfix1FilterForMatch()
4012 ** into spellfix1RunQuery().
4013 */
4014 typedef struct MatchQuery {
4015- spellfix1_cursor *pCur; /* The cursor being queried */
4016- sqlite3_stmt *pStmt; /* shadow table query statment */
4017- char zHash[SPELLFIX_MX_HASH]; /* The current phonehash for zPattern */
4018- const char *zPattern; /* Transliterated input string */
4019- int nPattern; /* Length of zPattern */
4020- EditDist3FromString *pMatchStr3; /* Original unicode string */
4021- EditDist3Config *pConfig3; /* Edit-distance cost coefficients */
4022- const EditDist3Lang *pLang; /* The selected language coefficients */
4023- int iLang; /* The language id */
4024- int iScope; /* Default scope */
4025- int iMaxDist; /* Maximum allowed edit distance, or -1 */
4026- int rc; /* Error code */
4027- int nRun; /* Number of prior runs for the same zPattern */
4028- char azPrior[SPELLFIX_MX_RUN][SPELLFIX_MX_HASH]; /* Prior hashes */
4029+ spellfix1_cursor *pCur; /* The cursor being queried */
4030+ sqlite3_stmt *pStmt; /* shadow table query statment */
4031+ char zHash[SPELLFIX_MX_HASH]; /* The current phonehash for zPattern */
4032+ const char *zPattern; /* Transliterated input string */
4033+ int nPattern; /* Length of zPattern */
4034+ EditDist3FromString *pMatchStr3; /* Original unicode string */
4035+ EditDist3Config *pConfig3; /* Edit-distance cost coefficients */
4036+ const EditDist3Lang *pLang; /* The selected language coefficients */
4037+ int iLang; /* The language id */
4038+ int iScope; /* Default scope */
4039+ int iMaxDist; /* Maximum allowed edit distance, or -1 */
4040+ int rc; /* Error code */
4041+ int nRun; /* Number of prior runs for the same zPattern */
4042+ char azPrior[SPELLFIX_MX_RUN][SPELLFIX_MX_HASH]; /* Prior hashes */
4043 } MatchQuery;
4044
4045 /*
4046 ** Run a query looking for the best matches against zPattern using
4047 ** zHash as the character class seed hash.
4048 */
4049-static void spellfix1RunQuery(MatchQuery *p, const char *zQuery, int nQuery){
4050- const char *zK1;
4051- const char *zWord;
4052- int iDist;
4053- int iRank;
4054- int iScore;
4055- int iWorst = 0;
4056- int idx;
4057- int idxWorst = -1;
4058- int i;
4059- int iScope = p->iScope;
4060- spellfix1_cursor *pCur = p->pCur;
4061- sqlite3_stmt *pStmt = p->pStmt;
4062- char zHash1[SPELLFIX_MX_HASH];
4063- char zHash2[SPELLFIX_MX_HASH];
4064- char *zClass;
4065- int nClass;
4066- int rc;
4067+static void spellfix1RunQuery(MatchQuery *p, const char *zQuery, int nQuery)
4068+{
4069+ const char *zK1;
4070+ const char *zWord;
4071+ int iDist;
4072+ int iRank;
4073+ int iScore;
4074+ int iWorst = 0;
4075+ int idx;
4076+ int idxWorst = -1;
4077+ int i;
4078+ int iScope = p->iScope;
4079+ spellfix1_cursor *pCur = p->pCur;
4080+ sqlite3_stmt *pStmt = p->pStmt;
4081+ char zHash1[SPELLFIX_MX_HASH];
4082+ char zHash2[SPELLFIX_MX_HASH];
4083+ char *zClass;
4084+ int nClass;
4085+ int rc;
4086
4087- if( pCur->a==0 || p->rc ) return; /* Prior memory allocation failure */
4088- zClass = (char*)phoneticHash((unsigned char*)zQuery, nQuery);
4089- if( zClass==0 ){
4090- p->rc = SQLITE_NOMEM;
4091- return;
4092- }
4093- nClass = (int)strlen(zClass);
4094- if( nClass>SPELLFIX_MX_HASH-2 ){
4095- nClass = SPELLFIX_MX_HASH-2;
4096- zClass[nClass] = 0;
4097- }
4098- if( nClass<=iScope ){
4099- if( nClass>2 ){
4100- iScope = nClass-1;
4101- }else{
4102- iScope = nClass;
4103+ if( pCur->a==0 || p->rc ) return; /* Prior memory allocation failure */
4104+ zClass = (char*)phoneticHash((unsigned char*)zQuery, nQuery);
4105+ if( zClass==0 ) {
4106+ p->rc = SQLITE_NOMEM;
4107+ return;
4108 }
4109- }
4110- memcpy(zHash1, zClass, iScope);
4111- sqlite3_free(zClass);
4112- zHash1[iScope] = 0;
4113- memcpy(zHash2, zHash1, iScope);
4114- zHash2[iScope] = 'Z';
4115- zHash2[iScope+1] = 0;
4116-#if SPELLFIX_MX_RUN>1
4117- for(i=0; i<p->nRun; i++){
4118- if( strcmp(p->azPrior[i], zHash1)==0 ) return;
4119- }
4120-#endif
4121- assert( p->nRun<SPELLFIX_MX_RUN );
4122- memcpy(p->azPrior[p->nRun++], zHash1, iScope+1);
4123- if( sqlite3_bind_text(pStmt, 1, zHash1, -1, SQLITE_STATIC)==SQLITE_NOMEM
4124- || sqlite3_bind_text(pStmt, 2, zHash2, -1, SQLITE_STATIC)==SQLITE_NOMEM
4125- ){
4126- p->rc = SQLITE_NOMEM;
4127- return;
4128- }
4129+ nClass = (int)strlen(zClass);
4130+ if( nClass>SPELLFIX_MX_HASH-2 ) {
4131+ nClass = SPELLFIX_MX_HASH-2;
4132+ zClass[nClass] = 0;
4133+ }
4134+ if( nClass<=iScope ) {
4135+ if( nClass>2 ) {
4136+ iScope = nClass-1;
4137+ } else {
4138+ iScope = nClass;
4139+ }
4140+ }
4141+ memcpy(zHash1, zClass, iScope);
4142+ sqlite3_free(zClass);
4143+ zHash1[iScope] = 0;
4144+ memcpy(zHash2, zHash1, iScope);
4145+ zHash2[iScope] = 'Z';
4146+ zHash2[iScope+1] = 0;
4147 #if SPELLFIX_MX_RUN>1
4148- for(i=0; i<pCur->nRow; i++){
4149- if( pCur->a[i].iScore>iWorst ){
4150- iWorst = pCur->a[i].iScore;
4151- idxWorst = i;
4152+ for(i=0; i<p->nRun; i++) {
4153+ if( strcmp(p->azPrior[i], zHash1)==0 ) return;
4154 }
4155- }
4156 #endif
4157- while( sqlite3_step(pStmt)==SQLITE_ROW ){
4158- int iMatchlen = -1;
4159- iRank = sqlite3_column_int(pStmt, 2);
4160- if( p->pMatchStr3 ){
4161- int nWord = sqlite3_column_bytes(pStmt, 1);
4162- zWord = (const char*)sqlite3_column_text(pStmt, 1);
4163- iDist = editDist3Core(p->pMatchStr3, zWord, nWord, p->pLang, &iMatchlen);
4164- }else{
4165- zK1 = (const char*)sqlite3_column_text(pStmt, 3);
4166- if( zK1==0 ) continue;
4167- iDist = editdist1(p->zPattern, zK1, 0);
4168- }
4169- if( iDist<0 ){
4170- p->rc = SQLITE_NOMEM;
4171- break;
4172+ assert( p->nRun<SPELLFIX_MX_RUN );
4173+ memcpy(p->azPrior[p->nRun++], zHash1, iScope+1);
4174+ if( sqlite3_bind_text(pStmt, 1, zHash1, -1, SQLITE_STATIC)==SQLITE_NOMEM
4175+ || sqlite3_bind_text(pStmt, 2, zHash2, -1, SQLITE_STATIC)==SQLITE_NOMEM
4176+ ) {
4177+ p->rc = SQLITE_NOMEM;
4178+ return;
4179 }
4180- pCur->nSearch++;
4181-
4182- /* If there is a "distance < $dist" or "distance <= $dist" constraint,
4183- ** check if this row meets it. If not, jump back up to the top of the
4184- ** loop to process the next row. Otherwise, if the row does match the
4185- ** distance constraint, check if the pCur->a[] array is already full.
4186- ** If it is and no explicit "top = ?" constraint was present in the
4187- ** query, grow the array to ensure there is room for the new entry. */
4188- assert( (p->iMaxDist>=0)==((pCur->idxNum & SPELLFIX_IDXNUM_DIST) ? 1 : 0) );
4189- if( p->iMaxDist>=0 ){
4190- if( iDist>p->iMaxDist ) continue;
4191- if( pCur->nRow>=pCur->nAlloc && (pCur->idxNum & SPELLFIX_IDXNUM_TOP)==0 ){
4192- spellfix1ResizeCursor(pCur, pCur->nAlloc*2 + 10);
4193- if( pCur->a==0 ) break;
4194- }
4195+#if SPELLFIX_MX_RUN>1
4196+ for(i=0; i<pCur->nRow; i++) {
4197+ if( pCur->a[i].iScore>iWorst ) {
4198+ iWorst = pCur->a[i].iScore;
4199+ idxWorst = i;
4200+ }
4201 }
4202+#endif
4203+ while( sqlite3_step(pStmt)==SQLITE_ROW ) {
4204+ int iMatchlen = -1;
4205+ iRank = sqlite3_column_int(pStmt, 2);
4206+ if( p->pMatchStr3 ) {
4207+ int nWord = sqlite3_column_bytes(pStmt, 1);
4208+ zWord = (const char*)sqlite3_column_text(pStmt, 1);
4209+ iDist = editDist3Core(p->pMatchStr3, zWord, nWord, p->pLang, &iMatchlen);
4210+ } else {
4211+ zK1 = (const char*)sqlite3_column_text(pStmt, 3);
4212+ if( zK1==0 ) continue;
4213+ iDist = editdist1(p->zPattern, zK1, 0);
4214+ }
4215+ if( iDist<0 ) {
4216+ p->rc = SQLITE_NOMEM;
4217+ break;
4218+ }
4219+ pCur->nSearch++;
4220
4221- iScore = spellfix1Score(iDist,iRank);
4222- if( pCur->nRow<pCur->nAlloc ){
4223- idx = pCur->nRow;
4224- }else if( iScore<iWorst ){
4225- idx = idxWorst;
4226- sqlite3_free(pCur->a[idx].zWord);
4227- }else{
4228- continue;
4229- }
4230+ /* If there is a "distance < $dist" or "distance <= $dist" constraint,
4231+ ** check if this row meets it. If not, jump back up to the top of the
4232+ ** loop to process the next row. Otherwise, if the row does match the
4233+ ** distance constraint, check if the pCur->a[] array is already full.
4234+ ** If it is and no explicit "top = ?" constraint was present in the
4235+ ** query, grow the array to ensure there is room for the new entry. */
4236+ assert( (p->iMaxDist>=0)==((pCur->idxNum & SPELLFIX_IDXNUM_DIST) ? 1 : 0) );
4237+ if( p->iMaxDist>=0 ) {
4238+ if( iDist>p->iMaxDist ) continue;
4239+ if( pCur->nRow>=pCur->nAlloc && (pCur->idxNum & SPELLFIX_IDXNUM_TOP)==0 ) {
4240+ spellfix1ResizeCursor(pCur, pCur->nAlloc*2 + 10);
4241+ if( pCur->a==0 ) break;
4242+ }
4243+ }
4244
4245- pCur->a[idx].zWord = sqlite3_mprintf("%s", sqlite3_column_text(pStmt, 1));
4246- if( pCur->a[idx].zWord==0 ){
4247- p->rc = SQLITE_NOMEM;
4248- break;
4249- }
4250- pCur->a[idx].iRowid = sqlite3_column_int64(pStmt, 0);
4251- pCur->a[idx].iRank = iRank;
4252- pCur->a[idx].iDistance = iDist;
4253- pCur->a[idx].iScore = iScore;
4254- pCur->a[idx].iMatchlen = iMatchlen;
4255- memcpy(pCur->a[idx].zHash, zHash1, iScope+1);
4256- if( pCur->nRow<pCur->nAlloc ) pCur->nRow++;
4257- if( pCur->nRow==pCur->nAlloc ){
4258- iWorst = pCur->a[0].iScore;
4259- idxWorst = 0;
4260- for(i=1; i<pCur->nRow; i++){
4261- iScore = pCur->a[i].iScore;
4262- if( iWorst<iScore ){
4263- iWorst = iScore;
4264- idxWorst = i;
4265+ iScore = spellfix1Score(iDist,iRank);
4266+ if( pCur->nRow<pCur->nAlloc ) {
4267+ idx = pCur->nRow;
4268+ } else if( iScore<iWorst ) {
4269+ idx = idxWorst;
4270+ sqlite3_free(pCur->a[idx].zWord);
4271+ } else {
4272+ continue;
4273 }
4274- }
4275+
4276+ pCur->a[idx].zWord = sqlite3_mprintf("%s", sqlite3_column_text(pStmt, 1));
4277+ if( pCur->a[idx].zWord==0 ) {
4278+ p->rc = SQLITE_NOMEM;
4279+ break;
4280+ }
4281+ pCur->a[idx].iRowid = sqlite3_column_int64(pStmt, 0);
4282+ pCur->a[idx].iRank = iRank;
4283+ pCur->a[idx].iDistance = iDist;
4284+ pCur->a[idx].iScore = iScore;
4285+ pCur->a[idx].iMatchlen = iMatchlen;
4286+ memcpy(pCur->a[idx].zHash, zHash1, iScope+1);
4287+ if( pCur->nRow<pCur->nAlloc ) pCur->nRow++;
4288+ if( pCur->nRow==pCur->nAlloc ) {
4289+ iWorst = pCur->a[0].iScore;
4290+ idxWorst = 0;
4291+ for(i=1; i<pCur->nRow; i++) {
4292+ iScore = pCur->a[i].iScore;
4293+ if( iWorst<iScore ) {
4294+ iWorst = iScore;
4295+ idxWorst = i;
4296+ }
4297+ }
4298+ }
4299 }
4300- }
4301- rc = sqlite3_reset(pStmt);
4302- if( rc ) p->rc = rc;
4303+ rc = sqlite3_reset(pStmt);
4304+ if( rc ) p->rc = rc;
4305 }
4306
4307 /*
4308@@ -2524,154 +2596,160 @@ ** This version of the xFilter method work if the MATCH term is present
4309 ** and we are doing a scan.
4310 */
4311 static int spellfix1FilterForMatch(
4312- spellfix1_cursor *pCur,
4313- int argc,
4314- sqlite3_value **argv
4315-){
4316- int idxNum = pCur->idxNum;
4317- const unsigned char *zMatchThis; /* RHS of the MATCH operator */
4318- EditDist3FromString *pMatchStr3 = 0; /* zMatchThis as an editdist string */
4319- char *zPattern; /* Transliteration of zMatchThis */
4320- int nPattern; /* Length of zPattern */
4321- int iLimit = 20; /* Max number of rows of output */
4322- int iScope = 3; /* Use this many characters of zClass */
4323- int iLang = 0; /* Language code */
4324- char *zSql; /* SQL of shadow table query */
4325- sqlite3_stmt *pStmt = 0; /* Shadow table query */
4326- int rc; /* Result code */
4327- int idx = 1; /* Next available filter parameter */
4328- spellfix1_vtab *p = pCur->pVTab; /* The virtual table that owns pCur */
4329- MatchQuery x; /* For passing info to RunQuery() */
4330+ spellfix1_cursor *pCur,
4331+ int argc,
4332+ sqlite3_value **argv
4333+)
4334+{
4335+ int idxNum = pCur->idxNum;
4336+ const unsigned char *zMatchThis; /* RHS of the MATCH operator */
4337+ EditDist3FromString *pMatchStr3 = 0; /* zMatchThis as an editdist string */
4338+ char *zPattern; /* Transliteration of zMatchThis */
4339+ int nPattern; /* Length of zPattern */
4340+ int iLimit = 20; /* Max number of rows of output */
4341+ int iScope = 3; /* Use this many characters of zClass */
4342+ int iLang = 0; /* Language code */
4343+ char *zSql; /* SQL of shadow table query */
4344+ sqlite3_stmt *pStmt = 0; /* Shadow table query */
4345+ int rc; /* Result code */
4346+ int idx = 1; /* Next available filter parameter */
4347+ spellfix1_vtab *p = pCur->pVTab; /* The virtual table that owns pCur */
4348+ MatchQuery x; /* For passing info to RunQuery() */
4349
4350- /* Load the cost table if we have not already done so */
4351- if( p->zCostTable!=0 && p->pConfig3==0 ){
4352- p->pConfig3 = sqlite3_malloc64( sizeof(p->pConfig3[0]) );
4353- if( p->pConfig3==0 ) return SQLITE_NOMEM;
4354- memset(p->pConfig3, 0, sizeof(p->pConfig3[0]));
4355- rc = editDist3ConfigLoad(p->pConfig3, p->db, p->zCostTable);
4356- if( rc ) return rc;
4357- }
4358- memset(&x, 0, sizeof(x));
4359- x.iScope = 3; /* Default scope if none specified by "WHERE scope=N" */
4360- x.iMaxDist = -1; /* Maximum allowed edit distance */
4361+ /* Load the cost table if we have not already done so */
4362+ if( p->zCostTable!=0 && p->pConfig3==0 ) {
4363+ p->pConfig3 = sqlite3_malloc64( sizeof(p->pConfig3[0]) );
4364+ if( p->pConfig3==0 ) return SQLITE_NOMEM;
4365+ memset(p->pConfig3, 0, sizeof(p->pConfig3[0]));
4366+ rc = editDist3ConfigLoad(p->pConfig3, p->db, p->zCostTable);
4367+ if( rc ) return rc;
4368+ }
4369+ memset(&x, 0, sizeof(x));
4370+ x.iScope = 3; /* Default scope if none specified by "WHERE scope=N" */
4371+ x.iMaxDist = -1; /* Maximum allowed edit distance */
4372
4373- if( idxNum&2 ){
4374- iLang = sqlite3_value_int(argv[idx++]);
4375- }
4376- if( idxNum&4 ){
4377- iLimit = sqlite3_value_int(argv[idx++]);
4378- if( iLimit<1 ) iLimit = 1;
4379- }
4380- if( idxNum&8 ){
4381- x.iScope = sqlite3_value_int(argv[idx++]);
4382- if( x.iScope<1 ) x.iScope = 1;
4383- if( x.iScope>SPELLFIX_MX_HASH-2 ) x.iScope = SPELLFIX_MX_HASH-2;
4384- }
4385- if( idxNum&(16|32) ){
4386- x.iMaxDist = sqlite3_value_int(argv[idx++]);
4387- if( idxNum&16 ) x.iMaxDist--;
4388- if( x.iMaxDist<0 ) x.iMaxDist = 0;
4389- }
4390- spellfix1ResetCursor(pCur);
4391- spellfix1ResizeCursor(pCur, iLimit);
4392- zMatchThis = sqlite3_value_text(argv[0]);
4393- if( zMatchThis==0 ) return SQLITE_OK;
4394- if( p->pConfig3 ){
4395- x.pLang = editDist3FindLang(p->pConfig3, iLang);
4396- pMatchStr3 = editDist3FromStringNew(x.pLang, (const char*)zMatchThis, -1);
4397- if( pMatchStr3==0 ){
4398- x.rc = SQLITE_NOMEM;
4399- goto filter_exit;
4400+ if( idxNum&2 ) {
4401+ iLang = sqlite3_value_int(argv[idx++]);
4402 }
4403- }else{
4404- x.pLang = 0;
4405- }
4406- zPattern = (char*)transliterate(zMatchThis, sqlite3_value_bytes(argv[0]));
4407- sqlite3_free(pCur->zPattern);
4408- pCur->zPattern = zPattern;
4409- if( zPattern==0 ){
4410- x.rc = SQLITE_NOMEM;
4411- goto filter_exit;
4412- }
4413- nPattern = (int)strlen(zPattern);
4414- if( zPattern[nPattern-1]=='*' ) nPattern--;
4415- zSql = sqlite3_mprintf(
4416- "SELECT id, word, rank, coalesce(k1,word)"
4417- " FROM \"%w\".\"%w_vocab\""
4418- " WHERE langid=%d AND k2>=?1 AND k2<?2",
4419- p->zDbName, p->zTableName, iLang
4420- );
4421- if( zSql==0 ){
4422- x.rc = SQLITE_NOMEM;
4423- pStmt = 0;
4424- goto filter_exit;
4425- }
4426- rc = sqlite3_prepare_v2(p->db, zSql, -1, &pStmt, 0);
4427- sqlite3_free(zSql);
4428- pCur->iLang = iLang;
4429- x.pCur = pCur;
4430- x.pStmt = pStmt;
4431- x.zPattern = zPattern;
4432- x.nPattern = nPattern;
4433- x.pMatchStr3 = pMatchStr3;
4434- x.iLang = iLang;
4435- x.rc = rc;
4436- x.pConfig3 = p->pConfig3;
4437- if( x.rc==SQLITE_OK ){
4438- spellfix1RunQuery(&x, zPattern, nPattern);
4439- }
4440+ if( idxNum&4 ) {
4441+ iLimit = sqlite3_value_int(argv[idx++]);
4442+ if( iLimit<1 ) iLimit = 1;
4443+ }
4444+ if( idxNum&8 ) {
4445+ x.iScope = sqlite3_value_int(argv[idx++]);
4446+ if( x.iScope<1 ) x.iScope = 1;
4447+ if( x.iScope>SPELLFIX_MX_HASH-2 ) x.iScope = SPELLFIX_MX_HASH-2;
4448+ }
4449+ if( idxNum&(16|32) ) {
4450+ x.iMaxDist = sqlite3_value_int(argv[idx++]);
4451+ if( idxNum&16 ) x.iMaxDist--;
4452+ if( x.iMaxDist<0 ) x.iMaxDist = 0;
4453+ }
4454+ spellfix1ResetCursor(pCur);
4455+ spellfix1ResizeCursor(pCur, iLimit);
4456+ zMatchThis = sqlite3_value_text(argv[0]);
4457+ if( zMatchThis==0 ) return SQLITE_OK;
4458+ if( p->pConfig3 ) {
4459+ x.pLang = editDist3FindLang(p->pConfig3, iLang);
4460+ pMatchStr3 = editDist3FromStringNew(x.pLang, (const char*)zMatchThis, -1);
4461+ if( pMatchStr3==0 ) {
4462+ x.rc = SQLITE_NOMEM;
4463+ goto filter_exit;
4464+ }
4465+ } else {
4466+ x.pLang = 0;
4467+ }
4468+ zPattern = (char*)transliterate(zMatchThis, sqlite3_value_bytes(argv[0]));
4469+ sqlite3_free(pCur->zPattern);
4470+ pCur->zPattern = zPattern;
4471+ if( zPattern==0 ) {
4472+ x.rc = SQLITE_NOMEM;
4473+ goto filter_exit;
4474+ }
4475+ nPattern = (int)strlen(zPattern);
4476+ if( zPattern[nPattern-1]=='*' ) nPattern--;
4477+ zSql = sqlite3_mprintf(
4478+ "SELECT id, word, rank, coalesce(k1,word)"
4479+ " FROM \"%w\".\"%w_vocab\""
4480+ " WHERE langid=%d AND k2>=?1 AND k2<?2",
4481+ p->zDbName, p->zTableName, iLang
4482+ );
4483+ if( zSql==0 ) {
4484+ x.rc = SQLITE_NOMEM;
4485+ pStmt = 0;
4486+ goto filter_exit;
4487+ }
4488+ rc = sqlite3_prepare_v2(p->db, zSql, -1, &pStmt, 0);
4489+ sqlite3_free(zSql);
4490+ pCur->iLang = iLang;
4491+ x.pCur = pCur;
4492+ x.pStmt = pStmt;
4493+ x.zPattern = zPattern;
4494+ x.nPattern = nPattern;
4495+ x.pMatchStr3 = pMatchStr3;
4496+ x.iLang = iLang;
4497+ x.rc = rc;
4498+ x.pConfig3 = p->pConfig3;
4499+ if( x.rc==SQLITE_OK ) {
4500+ spellfix1RunQuery(&x, zPattern, nPattern);
4501+ }
4502
4503- if( pCur->a ){
4504- qsort(pCur->a, pCur->nRow, sizeof(pCur->a[0]), spellfix1RowCompare);
4505- pCur->iTop = iLimit;
4506- pCur->iScope = iScope;
4507- }else{
4508- x.rc = SQLITE_NOMEM;
4509- }
4510+ if( pCur->a ) {
4511+ qsort(pCur->a, pCur->nRow, sizeof(pCur->a[0]), spellfix1RowCompare);
4512+ pCur->iTop = iLimit;
4513+ pCur->iScope = iScope;
4514+ } else {
4515+ x.rc = SQLITE_NOMEM;
4516+ }
4517
4518 filter_exit:
4519- sqlite3_finalize(pStmt);
4520- editDist3FromStringDelete(pMatchStr3);
4521- return x.rc;
4522+ sqlite3_finalize(pStmt);
4523+ editDist3FromStringDelete(pMatchStr3);
4524+ return x.rc;
4525 }
4526
4527 /*
4528 ** This version of xFilter handles a full-table scan case
4529 */
4530 static int spellfix1FilterForFullScan(
4531- spellfix1_cursor *pCur,
4532- int argc,
4533- sqlite3_value **argv
4534-){
4535- int rc = SQLITE_OK;
4536- int idxNum = pCur->idxNum;
4537- char *zSql;
4538- spellfix1_vtab *pVTab = pCur->pVTab;
4539- spellfix1ResetCursor(pCur);
4540- assert( idxNum==0 || idxNum==64 );
4541- zSql = sqlite3_mprintf(
4542- "SELECT word, rank, NULL, langid, id FROM \"%w\".\"%w_vocab\"%s",
4543- pVTab->zDbName, pVTab->zTableName,
4544- ((idxNum & 64) ? " WHERE rowid=?" : "")
4545- );
4546- if( zSql==0 ) return SQLITE_NOMEM;
4547- rc = sqlite3_prepare_v2(pVTab->db, zSql, -1, &pCur->pFullScan, 0);
4548- sqlite3_free(zSql);
4549- if( rc==SQLITE_OK && (idxNum & 64) ){
4550- assert( argc==1 );
4551- rc = sqlite3_bind_value(pCur->pFullScan, 1, argv[0]);
4552- }
4553- pCur->nRow = pCur->iRow = 0;
4554- if( rc==SQLITE_OK ){
4555- rc = sqlite3_step(pCur->pFullScan);
4556- if( rc==SQLITE_ROW ){ pCur->iRow = -1; rc = SQLITE_OK; }
4557- if( rc==SQLITE_DONE ){ rc = SQLITE_OK; }
4558- }else{
4559- pCur->iRow = 0;
4560- }
4561- return rc;
4562+ spellfix1_cursor *pCur,
4563+ int argc,
4564+ sqlite3_value **argv
4565+)
4566+{
4567+ int rc = SQLITE_OK;
4568+ int idxNum = pCur->idxNum;
4569+ char *zSql;
4570+ spellfix1_vtab *pVTab = pCur->pVTab;
4571+ spellfix1ResetCursor(pCur);
4572+ assert( idxNum==0 || idxNum==64 );
4573+ zSql = sqlite3_mprintf(
4574+ "SELECT word, rank, NULL, langid, id FROM \"%w\".\"%w_vocab\"%s",
4575+ pVTab->zDbName, pVTab->zTableName,
4576+ ((idxNum & 64) ? " WHERE rowid=?" : "")
4577+ );
4578+ if( zSql==0 ) return SQLITE_NOMEM;
4579+ rc = sqlite3_prepare_v2(pVTab->db, zSql, -1, &pCur->pFullScan, 0);
4580+ sqlite3_free(zSql);
4581+ if( rc==SQLITE_OK && (idxNum & 64) ) {
4582+ assert( argc==1 );
4583+ rc = sqlite3_bind_value(pCur->pFullScan, 1, argv[0]);
4584+ }
4585+ pCur->nRow = pCur->iRow = 0;
4586+ if( rc==SQLITE_OK ) {
4587+ rc = sqlite3_step(pCur->pFullScan);
4588+ if( rc==SQLITE_ROW ) {
4589+ pCur->iRow = -1;
4590+ rc = SQLITE_OK;
4591+ }
4592+ if( rc==SQLITE_DONE ) {
4593+ rc = SQLITE_OK;
4594+ }
4595+ } else {
4596+ pCur->iRow = 0;
4597+ }
4598+ return rc;
4599 }
4600-
4601
4602 /*
4603 ** Called to "rewind" a cursor back to the beginning so that
4604@@ -2679,145 +2757,149 @@ ** it starts its output over again. Always called at least once
4605 ** prior to any spellfix1Column, spellfix1Rowid, or spellfix1Eof call.
4606 */
4607 static int spellfix1Filter(
4608- sqlite3_vtab_cursor *cur,
4609- int idxNum, const char *idxStr,
4610- int argc, sqlite3_value **argv
4611-){
4612- spellfix1_cursor *pCur = (spellfix1_cursor *)cur;
4613- int rc;
4614- pCur->idxNum = idxNum;
4615- if( idxNum & 1 ){
4616- rc = spellfix1FilterForMatch(pCur, argc, argv);
4617- }else{
4618- rc = spellfix1FilterForFullScan(pCur, argc, argv);
4619- }
4620- return rc;
4621+ sqlite3_vtab_cursor *cur,
4622+ int idxNum, const char *idxStr,
4623+ int argc, sqlite3_value **argv
4624+)
4625+{
4626+ spellfix1_cursor *pCur = (spellfix1_cursor *)cur;
4627+ int rc;
4628+ pCur->idxNum = idxNum;
4629+ if( idxNum & 1 ) {
4630+ rc = spellfix1FilterForMatch(pCur, argc, argv);
4631+ } else {
4632+ rc = spellfix1FilterForFullScan(pCur, argc, argv);
4633+ }
4634+ return rc;
4635 }
4636
4637-
4638 /*
4639 ** Advance a cursor to its next row of output
4640 */
4641-static int spellfix1Next(sqlite3_vtab_cursor *cur){
4642- spellfix1_cursor *pCur = (spellfix1_cursor *)cur;
4643- int rc = SQLITE_OK;
4644- if( pCur->iRow < pCur->nRow ){
4645- if( pCur->pFullScan ){
4646- rc = sqlite3_step(pCur->pFullScan);
4647- if( rc!=SQLITE_ROW ) pCur->iRow = pCur->nRow;
4648- if( rc==SQLITE_ROW || rc==SQLITE_DONE ) rc = SQLITE_OK;
4649- }else{
4650- pCur->iRow++;
4651+static int spellfix1Next(sqlite3_vtab_cursor *cur)
4652+{
4653+ spellfix1_cursor *pCur = (spellfix1_cursor *)cur;
4654+ int rc = SQLITE_OK;
4655+ if( pCur->iRow < pCur->nRow ) {
4656+ if( pCur->pFullScan ) {
4657+ rc = sqlite3_step(pCur->pFullScan);
4658+ if( rc!=SQLITE_ROW ) pCur->iRow = pCur->nRow;
4659+ if( rc==SQLITE_ROW || rc==SQLITE_DONE ) rc = SQLITE_OK;
4660+ } else {
4661+ pCur->iRow++;
4662+ }
4663 }
4664- }
4665- return rc;
4666+ return rc;
4667 }
4668
4669 /*
4670 ** Return TRUE if we are at the end-of-file
4671 */
4672-static int spellfix1Eof(sqlite3_vtab_cursor *cur){
4673- spellfix1_cursor *pCur = (spellfix1_cursor *)cur;
4674- return pCur->iRow>=pCur->nRow;
4675+static int spellfix1Eof(sqlite3_vtab_cursor *cur)
4676+{
4677+ spellfix1_cursor *pCur = (spellfix1_cursor *)cur;
4678+ return pCur->iRow>=pCur->nRow;
4679 }
4680
4681 /*
4682 ** Return columns from the current row.
4683 */
4684 static int spellfix1Column(
4685- sqlite3_vtab_cursor *cur,
4686- sqlite3_context *ctx,
4687- int i
4688-){
4689- spellfix1_cursor *pCur = (spellfix1_cursor*)cur;
4690- if( pCur->pFullScan ){
4691- if( i<=SPELLFIX_COL_LANGID ){
4692- sqlite3_result_value(ctx, sqlite3_column_value(pCur->pFullScan, i));
4693- }else{
4694- sqlite3_result_null(ctx);
4695+ sqlite3_vtab_cursor *cur,
4696+ sqlite3_context *ctx,
4697+ int i
4698+)
4699+{
4700+ spellfix1_cursor *pCur = (spellfix1_cursor*)cur;
4701+ if( pCur->pFullScan ) {
4702+ if( i<=SPELLFIX_COL_LANGID ) {
4703+ sqlite3_result_value(ctx, sqlite3_column_value(pCur->pFullScan, i));
4704+ } else {
4705+ sqlite3_result_null(ctx);
4706+ }
4707+ return SQLITE_OK;
4708 }
4709- return SQLITE_OK;
4710- }
4711- switch( i ){
4712+ switch( i ) {
4713 case SPELLFIX_COL_WORD: {
4714- sqlite3_result_text(ctx, pCur->a[pCur->iRow].zWord, -1, SQLITE_STATIC);
4715- break;
4716+ sqlite3_result_text(ctx, pCur->a[pCur->iRow].zWord, -1, SQLITE_STATIC);
4717+ break;
4718 }
4719 case SPELLFIX_COL_RANK: {
4720- sqlite3_result_int(ctx, pCur->a[pCur->iRow].iRank);
4721- break;
4722+ sqlite3_result_int(ctx, pCur->a[pCur->iRow].iRank);
4723+ break;
4724 }
4725 case SPELLFIX_COL_DISTANCE: {
4726- sqlite3_result_int(ctx, pCur->a[pCur->iRow].iDistance);
4727- break;
4728+ sqlite3_result_int(ctx, pCur->a[pCur->iRow].iDistance);
4729+ break;
4730 }
4731 case SPELLFIX_COL_LANGID: {
4732- sqlite3_result_int(ctx, pCur->iLang);
4733- break;
4734+ sqlite3_result_int(ctx, pCur->iLang);
4735+ break;
4736 }
4737 case SPELLFIX_COL_SCORE: {
4738- sqlite3_result_int(ctx, pCur->a[pCur->iRow].iScore);
4739- break;
4740+ sqlite3_result_int(ctx, pCur->a[pCur->iRow].iScore);
4741+ break;
4742 }
4743 case SPELLFIX_COL_MATCHLEN: {
4744- int iMatchlen = pCur->a[pCur->iRow].iMatchlen;
4745- if( iMatchlen<0 ){
4746- int nPattern = (int)strlen(pCur->zPattern);
4747- char *zWord = pCur->a[pCur->iRow].zWord;
4748- int nWord = (int)strlen(zWord);
4749+ int iMatchlen = pCur->a[pCur->iRow].iMatchlen;
4750+ if( iMatchlen<0 ) {
4751+ int nPattern = (int)strlen(pCur->zPattern);
4752+ char *zWord = pCur->a[pCur->iRow].zWord;
4753+ int nWord = (int)strlen(zWord);
4754
4755- if( nPattern>0 && pCur->zPattern[nPattern-1]=='*' ){
4756- char *zTranslit;
4757- int res;
4758- zTranslit = (char *)transliterate((unsigned char *)zWord, nWord);
4759- if( !zTranslit ) return SQLITE_NOMEM;
4760- res = editdist1(pCur->zPattern, zTranslit, &iMatchlen);
4761- sqlite3_free(zTranslit);
4762- if( res<0 ) return SQLITE_NOMEM;
4763- iMatchlen = translen_to_charlen(zWord, nWord, iMatchlen);
4764- }else{
4765- iMatchlen = utf8Charlen(zWord, nWord);
4766+ if( nPattern>0 && pCur->zPattern[nPattern-1]=='*' ) {
4767+ char *zTranslit;
4768+ int res;
4769+ zTranslit = (char *)transliterate((unsigned char *)zWord, nWord);
4770+ if( !zTranslit ) return SQLITE_NOMEM;
4771+ res = editdist1(pCur->zPattern, zTranslit, &iMatchlen);
4772+ sqlite3_free(zTranslit);
4773+ if( res<0 ) return SQLITE_NOMEM;
4774+ iMatchlen = translen_to_charlen(zWord, nWord, iMatchlen);
4775+ } else {
4776+ iMatchlen = utf8Charlen(zWord, nWord);
4777+ }
4778 }
4779- }
4780
4781- sqlite3_result_int(ctx, iMatchlen);
4782- break;
4783+ sqlite3_result_int(ctx, iMatchlen);
4784+ break;
4785 }
4786 case SPELLFIX_COL_PHONEHASH: {
4787- sqlite3_result_text(ctx, pCur->a[pCur->iRow].zHash, -1, SQLITE_STATIC);
4788- break;
4789+ sqlite3_result_text(ctx, pCur->a[pCur->iRow].zHash, -1, SQLITE_STATIC);
4790+ break;
4791 }
4792 case SPELLFIX_COL_TOP: {
4793- sqlite3_result_int(ctx, pCur->iTop);
4794- break;
4795+ sqlite3_result_int(ctx, pCur->iTop);
4796+ break;
4797 }
4798 case SPELLFIX_COL_SCOPE: {
4799- sqlite3_result_int(ctx, pCur->iScope);
4800- break;
4801+ sqlite3_result_int(ctx, pCur->iScope);
4802+ break;
4803 }
4804 case SPELLFIX_COL_SRCHCNT: {
4805- sqlite3_result_int(ctx, pCur->nSearch);
4806- break;
4807+ sqlite3_result_int(ctx, pCur->nSearch);
4808+ break;
4809 }
4810 default: {
4811- sqlite3_result_null(ctx);
4812- break;
4813+ sqlite3_result_null(ctx);
4814+ break;
4815+ }
4816 }
4817- }
4818- return SQLITE_OK;
4819+ return SQLITE_OK;
4820 }
4821
4822 /*
4823 ** The rowid.
4824 */
4825-static int spellfix1Rowid(sqlite3_vtab_cursor *cur, sqlite_int64 *pRowid){
4826- spellfix1_cursor *pCur = (spellfix1_cursor*)cur;
4827- if( pCur->pFullScan ){
4828- *pRowid = sqlite3_column_int64(pCur->pFullScan, 4);
4829- }else{
4830- *pRowid = pCur->a[pCur->iRow].iRowid;
4831- }
4832- return SQLITE_OK;
4833+static int spellfix1Rowid(sqlite3_vtab_cursor *cur, sqlite_int64 *pRowid)
4834+{
4835+ spellfix1_cursor *pCur = (spellfix1_cursor*)cur;
4836+ if( pCur->pFullScan ) {
4837+ *pRowid = sqlite3_column_int64(pCur->pFullScan, 4);
4838+ } else {
4839+ *pRowid = pCur->a[pCur->iRow].iRowid;
4840+ }
4841+ return SQLITE_OK;
4842 }
4843
4844 /*
4845@@ -2825,234 +2907,237 @@ ** This function is called by the xUpdate() method. It returns a string
4846 ** containing the conflict mode that xUpdate() should use for the current
4847 ** operation. One of: "ROLLBACK", "IGNORE", "ABORT" or "REPLACE".
4848 */
4849-static const char *spellfix1GetConflict(sqlite3 *db){
4850- static const char *azConflict[] = {
4851- /* Note: Instead of "FAIL" - "ABORT". */
4852- "ROLLBACK", "IGNORE", "ABORT", "ABORT", "REPLACE"
4853- };
4854- int eConflict = sqlite3_vtab_on_conflict(db);
4855+static const char *spellfix1GetConflict(sqlite3 *db)
4856+{
4857+ static const char *azConflict[] = {
4858+ /* Note: Instead of "FAIL" - "ABORT". */
4859+ "ROLLBACK", "IGNORE", "ABORT", "ABORT", "REPLACE"
4860+ };
4861+ int eConflict = sqlite3_vtab_on_conflict(db);
4862
4863- assert( eConflict==SQLITE_ROLLBACK || eConflict==SQLITE_IGNORE
4864- || eConflict==SQLITE_FAIL || eConflict==SQLITE_ABORT
4865- || eConflict==SQLITE_REPLACE
4866- );
4867- assert( SQLITE_ROLLBACK==1 );
4868- assert( SQLITE_IGNORE==2 );
4869- assert( SQLITE_FAIL==3 );
4870- assert( SQLITE_ABORT==4 );
4871- assert( SQLITE_REPLACE==5 );
4872+ assert( eConflict==SQLITE_ROLLBACK || eConflict==SQLITE_IGNORE
4873+ || eConflict==SQLITE_FAIL || eConflict==SQLITE_ABORT
4874+ || eConflict==SQLITE_REPLACE
4875+ );
4876+ assert( SQLITE_ROLLBACK==1 );
4877+ assert( SQLITE_IGNORE==2 );
4878+ assert( SQLITE_FAIL==3 );
4879+ assert( SQLITE_ABORT==4 );
4880+ assert( SQLITE_REPLACE==5 );
4881
4882- return azConflict[eConflict-1];
4883+ return azConflict[eConflict-1];
4884 }
4885
4886 /*
4887 ** The xUpdate() method.
4888 */
4889 static int spellfix1Update(
4890- sqlite3_vtab *pVTab,
4891- int argc,
4892- sqlite3_value **argv,
4893- sqlite_int64 *pRowid
4894-){
4895- int rc = SQLITE_OK;
4896- sqlite3_int64 rowid, newRowid;
4897- spellfix1_vtab *p = (spellfix1_vtab*)pVTab;
4898- sqlite3 *db = p->db;
4899+ sqlite3_vtab *pVTab,
4900+ int argc,
4901+ sqlite3_value **argv,
4902+ sqlite_int64 *pRowid
4903+)
4904+{
4905+ int rc = SQLITE_OK;
4906+ sqlite3_int64 rowid, newRowid;
4907+ spellfix1_vtab *p = (spellfix1_vtab*)pVTab;
4908+ sqlite3 *db = p->db;
4909
4910- if( argc==1 ){
4911- /* A delete operation on the rowid given by argv[0] */
4912- rowid = *pRowid = sqlite3_value_int64(argv[0]);
4913- spellfix1DbExec(&rc, db, "DELETE FROM \"%w\".\"%w_vocab\" "
4914- " WHERE id=%lld",
4915- p->zDbName, p->zTableName, rowid);
4916- }else{
4917- const unsigned char *zWord = sqlite3_value_text(argv[SPELLFIX_COL_WORD+2]);
4918- int nWord = sqlite3_value_bytes(argv[SPELLFIX_COL_WORD+2]);
4919- int iLang = sqlite3_value_int(argv[SPELLFIX_COL_LANGID+2]);
4920- int iRank = sqlite3_value_int(argv[SPELLFIX_COL_RANK+2]);
4921- const unsigned char *zSoundslike =
4922- sqlite3_value_text(argv[SPELLFIX_COL_SOUNDSLIKE+2]);
4923- int nSoundslike = sqlite3_value_bytes(argv[SPELLFIX_COL_SOUNDSLIKE+2]);
4924- char *zK1, *zK2;
4925- int i;
4926- char c;
4927- const char *zConflict = spellfix1GetConflict(db);
4928+ if( argc==1 ) {
4929+ /* A delete operation on the rowid given by argv[0] */
4930+ rowid = *pRowid = sqlite3_value_int64(argv[0]);
4931+ spellfix1DbExec(&rc, db, "DELETE FROM \"%w\".\"%w_vocab\" "
4932+ " WHERE id=%lld",
4933+ p->zDbName, p->zTableName, rowid);
4934+ } else {
4935+ const unsigned char *zWord = sqlite3_value_text(argv[SPELLFIX_COL_WORD+2]);
4936+ int nWord = sqlite3_value_bytes(argv[SPELLFIX_COL_WORD+2]);
4937+ int iLang = sqlite3_value_int(argv[SPELLFIX_COL_LANGID+2]);
4938+ int iRank = sqlite3_value_int(argv[SPELLFIX_COL_RANK+2]);
4939+ const unsigned char *zSoundslike =
4940+ sqlite3_value_text(argv[SPELLFIX_COL_SOUNDSLIKE+2]);
4941+ int nSoundslike = sqlite3_value_bytes(argv[SPELLFIX_COL_SOUNDSLIKE+2]);
4942+ char *zK1, *zK2;
4943+ int i;
4944+ char c;
4945+ const char *zConflict = spellfix1GetConflict(db);
4946
4947- if( zWord==0 ){
4948- /* Inserts of the form: INSERT INTO table(command) VALUES('xyzzy');
4949- ** cause zWord to be NULL, so we look at the "command" column to see
4950- ** what special actions to take */
4951- const char *zCmd =
4952- (const char*)sqlite3_value_text(argv[SPELLFIX_COL_COMMAND+2]);
4953- if( zCmd==0 ){
4954- pVTab->zErrMsg = sqlite3_mprintf("NOT NULL constraint failed: %s.word",
4955- p->zTableName);
4956- return SQLITE_CONSTRAINT_NOTNULL;
4957- }
4958- if( strcmp(zCmd,"reset")==0 ){
4959- /* Reset the edit cost table (if there is one). */
4960- editDist3ConfigDelete(p->pConfig3);
4961- p->pConfig3 = 0;
4962- return SQLITE_OK;
4963- }
4964- if( strncmp(zCmd,"edit_cost_table=",16)==0 ){
4965- editDist3ConfigDelete(p->pConfig3);
4966- p->pConfig3 = 0;
4967- sqlite3_free(p->zCostTable);
4968- p->zCostTable = spellfix1Dequote(zCmd+16);
4969- if( p->zCostTable==0 ) return SQLITE_NOMEM;
4970- if( p->zCostTable[0]==0 || sqlite3_stricmp(p->zCostTable,"null")==0 ){
4971- sqlite3_free(p->zCostTable);
4972- p->zCostTable = 0;
4973+ if( zWord==0 ) {
4974+ /* Inserts of the form: INSERT INTO table(command) VALUES('xyzzy');
4975+ ** cause zWord to be NULL, so we look at the "command" column to see
4976+ ** what special actions to take */
4977+ const char *zCmd =
4978+ (const char*)sqlite3_value_text(argv[SPELLFIX_COL_COMMAND+2]);
4979+ if( zCmd==0 ) {
4980+ pVTab->zErrMsg = sqlite3_mprintf("NOT NULL constraint failed: %s.word",
4981+ p->zTableName);
4982+ return SQLITE_CONSTRAINT_NOTNULL;
4983+ }
4984+ if( strcmp(zCmd,"reset")==0 ) {
4985+ /* Reset the edit cost table (if there is one). */
4986+ editDist3ConfigDelete(p->pConfig3);
4987+ p->pConfig3 = 0;
4988+ return SQLITE_OK;
4989+ }
4990+ if( strncmp(zCmd,"edit_cost_table=",16)==0 ) {
4991+ editDist3ConfigDelete(p->pConfig3);
4992+ p->pConfig3 = 0;
4993+ sqlite3_free(p->zCostTable);
4994+ p->zCostTable = spellfix1Dequote(zCmd+16);
4995+ if( p->zCostTable==0 ) return SQLITE_NOMEM;
4996+ if( p->zCostTable[0]==0 || sqlite3_stricmp(p->zCostTable,"null")==0 ) {
4997+ sqlite3_free(p->zCostTable);
4998+ p->zCostTable = 0;
4999+ }
5000+ return SQLITE_OK;
5001+ }
5002+ pVTab->zErrMsg = sqlite3_mprintf("unknown value for %s.command: \"%w\"",
5003+ p->zTableName, zCmd);
5004+ return SQLITE_ERROR;
5005 }
5006- return SQLITE_OK;
5007- }
5008- pVTab->zErrMsg = sqlite3_mprintf("unknown value for %s.command: \"%w\"",
5009- p->zTableName, zCmd);
5010- return SQLITE_ERROR;
5011- }
5012- if( iRank<1 ) iRank = 1;
5013- if( zSoundslike ){
5014- zK1 = (char*)transliterate(zSoundslike, nSoundslike);
5015- }else{
5016- zK1 = (char*)transliterate(zWord, nWord);
5017- }
5018- if( zK1==0 ) return SQLITE_NOMEM;
5019- for(i=0; (c = zK1[i])!=0; i++){
5020- if( c>='A' && c<='Z' ) zK1[i] += 'a' - 'A';
5021- }
5022- zK2 = (char*)phoneticHash((const unsigned char*)zK1, i);
5023- if( zK2==0 ){
5024- sqlite3_free(zK1);
5025- return SQLITE_NOMEM;
5026- }
5027- if( sqlite3_value_type(argv[0])==SQLITE_NULL ){
5028- if( sqlite3_value_type(argv[1])==SQLITE_NULL ){
5029- spellfix1DbExec(&rc, db,
5030- "INSERT INTO \"%w\".\"%w_vocab\"(rank,langid,word,k1,k2) "
5031- "VALUES(%d,%d,%Q,nullif(%Q,%Q),%Q)",
5032- p->zDbName, p->zTableName,
5033- iRank, iLang, zWord, zK1, zWord, zK2
5034- );
5035- }else{
5036- newRowid = sqlite3_value_int64(argv[1]);
5037- spellfix1DbExec(&rc, db,
5038- "INSERT OR %s INTO \"%w\".\"%w_vocab\"(id,rank,langid,word,k1,k2) "
5039- "VALUES(%lld,%d,%d,%Q,nullif(%Q,%Q),%Q)",
5040- zConflict, p->zDbName, p->zTableName,
5041- newRowid, iRank, iLang, zWord, zK1, zWord, zK2
5042- );
5043- }
5044- *pRowid = sqlite3_last_insert_rowid(db);
5045- }else{
5046- rowid = sqlite3_value_int64(argv[0]);
5047- newRowid = *pRowid = sqlite3_value_int64(argv[1]);
5048- spellfix1DbExec(&rc, db,
5049- "UPDATE OR %s \"%w\".\"%w_vocab\" SET id=%lld, rank=%d, langid=%d,"
5050- " word=%Q, k1=nullif(%Q,%Q), k2=%Q WHERE id=%lld",
5051- zConflict, p->zDbName, p->zTableName, newRowid, iRank, iLang,
5052- zWord, zK1, zWord, zK2, rowid
5053- );
5054+ if( iRank<1 ) iRank = 1;
5055+ if( zSoundslike ) {
5056+ zK1 = (char*)transliterate(zSoundslike, nSoundslike);
5057+ } else {
5058+ zK1 = (char*)transliterate(zWord, nWord);
5059+ }
5060+ if( zK1==0 ) return SQLITE_NOMEM;
5061+ for(i=0; (c = zK1[i])!=0; i++) {
5062+ if( c>='A' && c<='Z' ) zK1[i] += 'a' - 'A';
5063+ }
5064+ zK2 = (char*)phoneticHash((const unsigned char*)zK1, i);
5065+ if( zK2==0 ) {
5066+ sqlite3_free(zK1);
5067+ return SQLITE_NOMEM;
5068+ }
5069+ if( sqlite3_value_type(argv[0])==SQLITE_NULL ) {
5070+ if( sqlite3_value_type(argv[1])==SQLITE_NULL ) {
5071+ spellfix1DbExec(&rc, db,
5072+ "INSERT INTO \"%w\".\"%w_vocab\"(rank,langid,word,k1,k2) "
5073+ "VALUES(%d,%d,%Q,nullif(%Q,%Q),%Q)",
5074+ p->zDbName, p->zTableName,
5075+ iRank, iLang, zWord, zK1, zWord, zK2
5076+ );
5077+ } else {
5078+ newRowid = sqlite3_value_int64(argv[1]);
5079+ spellfix1DbExec(&rc, db,
5080+ "INSERT OR %s INTO \"%w\".\"%w_vocab\"(id,rank,langid,word,k1,k2) "
5081+ "VALUES(%lld,%d,%d,%Q,nullif(%Q,%Q),%Q)",
5082+ zConflict, p->zDbName, p->zTableName,
5083+ newRowid, iRank, iLang, zWord, zK1, zWord, zK2
5084+ );
5085+ }
5086+ *pRowid = sqlite3_last_insert_rowid(db);
5087+ } else {
5088+ rowid = sqlite3_value_int64(argv[0]);
5089+ newRowid = *pRowid = sqlite3_value_int64(argv[1]);
5090+ spellfix1DbExec(&rc, db,
5091+ "UPDATE OR %s \"%w\".\"%w_vocab\" SET id=%lld, rank=%d, langid=%d,"
5092+ " word=%Q, k1=nullif(%Q,%Q), k2=%Q WHERE id=%lld",
5093+ zConflict, p->zDbName, p->zTableName, newRowid, iRank, iLang,
5094+ zWord, zK1, zWord, zK2, rowid
5095+ );
5096+ }
5097+ sqlite3_free(zK1);
5098+ sqlite3_free(zK2);
5099 }
5100- sqlite3_free(zK1);
5101- sqlite3_free(zK2);
5102- }
5103- return rc;
5104+ return rc;
5105 }
5106
5107 /*
5108 ** Rename the spellfix1 table.
5109 */
5110-static int spellfix1Rename(sqlite3_vtab *pVTab, const char *zNew){
5111- spellfix1_vtab *p = (spellfix1_vtab*)pVTab;
5112- sqlite3 *db = p->db;
5113- int rc = SQLITE_OK;
5114- char *zNewName = sqlite3_mprintf("%s", zNew);
5115- if( zNewName==0 ){
5116- return SQLITE_NOMEM;
5117- }
5118- spellfix1DbExec(&rc, db,
5119- "ALTER TABLE \"%w\".\"%w_vocab\" RENAME TO \"%w_vocab\"",
5120- p->zDbName, p->zTableName, zNewName
5121- );
5122- if( rc==SQLITE_OK ){
5123- sqlite3_free(p->zTableName);
5124- p->zTableName = zNewName;
5125- }else{
5126- sqlite3_free(zNewName);
5127- }
5128- return rc;
5129+static int spellfix1Rename(sqlite3_vtab *pVTab, const char *zNew)
5130+{
5131+ spellfix1_vtab *p = (spellfix1_vtab*)pVTab;
5132+ sqlite3 *db = p->db;
5133+ int rc = SQLITE_OK;
5134+ char *zNewName = sqlite3_mprintf("%s", zNew);
5135+ if( zNewName==0 ) {
5136+ return SQLITE_NOMEM;
5137+ }
5138+ spellfix1DbExec(&rc, db,
5139+ "ALTER TABLE \"%w\".\"%w_vocab\" RENAME TO \"%w_vocab\"",
5140+ p->zDbName, p->zTableName, zNewName
5141+ );
5142+ if( rc==SQLITE_OK ) {
5143+ sqlite3_free(p->zTableName);
5144+ p->zTableName = zNewName;
5145+ } else {
5146+ sqlite3_free(zNewName);
5147+ }
5148+ return rc;
5149 }
5150
5151-
5152 /*
5153 ** A virtual table module that provides fuzzy search.
5154 */
5155 static sqlite3_module spellfix1Module = {
5156- 0, /* iVersion */
5157- spellfix1Create, /* xCreate - handle CREATE VIRTUAL TABLE */
5158- spellfix1Connect, /* xConnect - reconnected to an existing table */
5159- spellfix1BestIndex, /* xBestIndex - figure out how to do a query */
5160- spellfix1Disconnect, /* xDisconnect - close a connection */
5161- spellfix1Destroy, /* xDestroy - handle DROP TABLE */
5162- spellfix1Open, /* xOpen - open a cursor */
5163- spellfix1Close, /* xClose - close a cursor */
5164- spellfix1Filter, /* xFilter - configure scan constraints */
5165- spellfix1Next, /* xNext - advance a cursor */
5166- spellfix1Eof, /* xEof - check for end of scan */
5167- spellfix1Column, /* xColumn - read data */
5168- spellfix1Rowid, /* xRowid - read data */
5169- spellfix1Update, /* xUpdate */
5170- 0, /* xBegin */
5171- 0, /* xSync */
5172- 0, /* xCommit */
5173- 0, /* xRollback */
5174- 0, /* xFindMethod */
5175- spellfix1Rename, /* xRename */
5176- 0, /* xSavepoint */
5177- 0, /* xRelease */
5178- 0, /* xRollbackTo */
5179- 0, /* xShadowName */
5180- 0 /* xIntegrity */
5181+ 0, /* iVersion */
5182+ spellfix1Create, /* xCreate - handle CREATE VIRTUAL TABLE */
5183+ spellfix1Connect, /* xConnect - reconnected to an existing table */
5184+ spellfix1BestIndex, /* xBestIndex - figure out how to do a query */
5185+ spellfix1Disconnect, /* xDisconnect - close a connection */
5186+ spellfix1Destroy, /* xDestroy - handle DROP TABLE */
5187+ spellfix1Open, /* xOpen - open a cursor */
5188+ spellfix1Close, /* xClose - close a cursor */
5189+ spellfix1Filter, /* xFilter - configure scan constraints */
5190+ spellfix1Next, /* xNext - advance a cursor */
5191+ spellfix1Eof, /* xEof - check for end of scan */
5192+ spellfix1Column, /* xColumn - read data */
5193+ spellfix1Rowid, /* xRowid - read data */
5194+ spellfix1Update, /* xUpdate */
5195+ 0, /* xBegin */
5196+ 0, /* xSync */
5197+ 0, /* xCommit */
5198+ 0, /* xRollback */
5199+ 0, /* xFindMethod */
5200+ spellfix1Rename, /* xRename */
5201+ 0, /* xSavepoint */
5202+ 0, /* xRelease */
5203+ 0, /* xRollbackTo */
5204+ 0, /* xShadowName */
5205+ 0 /* xIntegrity */
5206 };
5207
5208 /*
5209 ** Register the various functions and the virtual table.
5210 */
5211-static int spellfix1Register(sqlite3 *db){
5212- int rc = SQLITE_OK;
5213- int i;
5214- rc = sqlite3_create_function(db, "spellfix1_translit", 1,
5215- SQLITE_UTF8|SQLITE_DETERMINISTIC, 0,
5216- transliterateSqlFunc, 0, 0);
5217- if( rc==SQLITE_OK ){
5218- rc = sqlite3_create_function(db, "spellfix1_editdist", 2,
5219+static int spellfix1Register(sqlite3 *db)
5220+{
5221+ int rc = SQLITE_OK;
5222+ int i;
5223+ rc = sqlite3_create_function(db, "spellfix1_translit", 1,
5224 SQLITE_UTF8|SQLITE_DETERMINISTIC, 0,
5225- editdistSqlFunc, 0, 0);
5226- }
5227- if( rc==SQLITE_OK ){
5228- rc = sqlite3_create_function(db, "spellfix1_phonehash", 1,
5229- SQLITE_UTF8|SQLITE_DETERMINISTIC, 0,
5230- phoneticHashSqlFunc, 0, 0);
5231- }
5232- if( rc==SQLITE_OK ){
5233- rc = sqlite3_create_function(db, "spellfix1_scriptcode", 1,
5234- SQLITE_UTF8|SQLITE_DETERMINISTIC, 0,
5235- scriptCodeSqlFunc, 0, 0);
5236- }
5237- if( rc==SQLITE_OK ){
5238- rc = sqlite3_create_module(db, "spellfix1", &spellfix1Module, 0);
5239- }
5240- if( rc==SQLITE_OK ){
5241- rc = editDist3Install(db);
5242- }
5243+ transliterateSqlFunc, 0, 0);
5244+ if( rc==SQLITE_OK ) {
5245+ rc = sqlite3_create_function(db, "spellfix1_editdist", 2,
5246+ SQLITE_UTF8|SQLITE_DETERMINISTIC, 0,
5247+ editdistSqlFunc, 0, 0);
5248+ }
5249+ if( rc==SQLITE_OK ) {
5250+ rc = sqlite3_create_function(db, "spellfix1_phonehash", 1,
5251+ SQLITE_UTF8|SQLITE_DETERMINISTIC, 0,
5252+ phoneticHashSqlFunc, 0, 0);
5253+ }
5254+ if( rc==SQLITE_OK ) {
5255+ rc = sqlite3_create_function(db, "spellfix1_scriptcode", 1,
5256+ SQLITE_UTF8|SQLITE_DETERMINISTIC, 0,
5257+ scriptCodeSqlFunc, 0, 0);
5258+ }
5259+ if( rc==SQLITE_OK ) {
5260+ rc = sqlite3_create_module(db, "spellfix1", &spellfix1Module, 0);
5261+ }
5262+ if( rc==SQLITE_OK ) {
5263+ rc = editDist3Install(db);
5264+ }
5265
5266- /* Verify sanity of the translit[] table */
5267- for(i=0; i<sizeof(translit)/sizeof(translit[0])-1; i++){
5268- assert( translit[i].cFrom<translit[i+1].cFrom );
5269- }
5270+ /* Verify sanity of the translit[] table */
5271+ for(i=0; i<sizeof(translit)/sizeof(translit[0])-1; i++) {
5272+ assert( translit[i].cFrom<translit[i+1].cFrom );
5273+ }
5274
5275- return rc;
5276+ return rc;
5277 }
5278
5279 #endif /* SQLITE_OMIT_VIRTUALTABLE */
5280@@ -3064,13 +3149,14 @@ #ifdef _WIN32
5281 __declspec(dllexport)
5282 #endif
5283 int sqlite3_spellfix_init(
5284- sqlite3 *db,
5285- char **pzErrMsg,
5286- const sqlite3_api_routines *pApi
5287-){
5288- SQLITE_EXTENSION_INIT2(pApi);
5289+ sqlite3 *db,
5290+ char **pzErrMsg,
5291+ const sqlite3_api_routines *pApi
5292+)
5293+{
5294+ SQLITE_EXTENSION_INIT2(pApi);
5295 #ifndef SQLITE_OMIT_VIRTUALTABLE
5296- return spellfix1Register(db);
5297+ return spellfix1Register(db);
5298 #endif
5299- return SQLITE_OK;
5300+ return SQLITE_OK;
5301 }
5302diff --git a/importer/main.c b/importer/main.c
5303index d03774569333e62028a03a3789df97878026daf6..b83e85086c5fac3b57c0fb27dfd4e7fd041282c8 100644
5304--- a/importer/main.c
5305+++ b/importer/main.c
5306@@ -7,27 +7,27 @@ #include "../lib/data.h"
5307
5308 int run(const char *db, const char *txt);
5309
5310-int main(int argc, char** argv) {
5311+int main(int argc, char** argv)
5312+{
5313 int opt;
5314 char* txt = NULL;
5315 char* db = NULL;
5316
5317 while ((opt = getopt(argc, argv, "t:d:h")) != -1) {
5318 switch(opt) {
5319- case 't':
5320- txt = copy_achar(optarg);
5321- break;
5322- case 'd':
5323- db = copy_achar(optarg);
5324- break;
5325- case 'h':
5326- // fall through
5327- default:
5328- printf("Usage: %s", argv[0]);
5329- goto end;
5330+ case 't':
5331+ txt = copy_achar(optarg);
5332+ break;
5333+ case 'd':
5334+ db = copy_achar(optarg);
5335+ break;
5336+ case 'h':
5337+ // fall through
5338+ default:
5339+ printf("Usage: %s", argv[0]);
5340+ goto end;
5341 }
5342 }
5343-
5344
5345 int r = run(db, txt);
5346
5347@@ -40,7 +40,8 @@
5348 return r;
5349 }
5350
5351-int run(const char *db, const char *txt) {
5352+int run(const char *db, const char *txt)
5353+{
5354 char * line = NULL;
5355 size_t len = 0;
5356 int count = 0;
5357@@ -66,7 +67,7 @@
5358 insert(data, line, read-1);
5359 count++;
5360
5361- if ((count % 321) == 0){
5362+ if ((count % 321) == 0) {
5363 float t = ((float)count/(float)total)*100;
5364 printf("\rLoading data [%03.0f%%] %d/%d", t, count, total);
5365 }
5366diff --git a/lib/data.c b/lib/data.c
5367index 7ebf597ca8ab317c76ac1960882f19669994d18e..e94e0a777c5de6b40530247fecb6942c985e803d 100644
5368--- a/lib/data.c
5369+++ b/lib/data.c
5370@@ -8,7 +8,8 @@ const char *insert_into = "INSERT INTO words (LINE) VALUES($VVV);";
5371 const char *select_words = "SELECT Id, Line FROM words WHERE line like $VVV LIMIT 10;";
5372 const char *create_table = "CREATE TABLE IF NOT EXISTS words (ID INTEGER PRIMARY KEY AUTOINCREMENT, LINE TEXT NOT NULL);";
5373
5374-Data* new_data(const char* con) {
5375+Data* new_data(const char* con)
5376+{
5377 Data* data = (Data*)malloc(sizeof(Data));
5378
5379 int v = sqlite3_open(con, &(data->db));
5380@@ -27,13 +28,15 @@
5381 return data;
5382 }
5383
5384-void free_data(Data* data) {
5385+void free_data(Data* data)
5386+{
5387 sqlite3_close(data->db);
5388 free(data);
5389 }
5390
5391-void insert(Data* data, char* line, int len) {
5392- sqlite3_stmt *stmt;
5393+void insert(Data* data, char* line, int len)
5394+{
5395+ sqlite3_stmt *stmt;
5396 int r = sqlite3_prepare_v2(data->db, insert_into, -1, &stmt, NULL);
5397
5398 if (r != SQLITE_OK) {
5399@@ -55,8 +58,9 @@
5400 sqlite3_finalize(stmt);
5401 }
5402
5403-void bootstrap(Data* data) {
5404- sqlite3_stmt *stmt;
5405+void bootstrap(Data* data)
5406+{
5407+ sqlite3_stmt *stmt;
5408 int r = sqlite3_prepare_v2(data->db, create_table, -1, &stmt, NULL);
5409
5410 if (r != SQLITE_OK) {
5411@@ -76,8 +80,9 @@
5412 sqlite3_finalize(stmt);
5413 }
5414
5415-LIST* data_select(Data* data, char *sch, int len) {
5416- sqlite3_stmt *stmt;
5417+LIST* data_select(Data* data, char *sch, int len)
5418+{
5419+ sqlite3_stmt *stmt;
5420 int r = sqlite3_prepare_v2(data->db, select_words, -1, &stmt, NULL);
5421
5422 if (r != SQLITE_OK) {
5423@@ -112,6 +117,7 @@
5424 return list;
5425 }
5426
5427-void print_result_code(int code) {
5428+void print_result_code(int code)
5429+{
5430 printf(sqlite3_errstr(code));
5431 }
5432diff --git a/lib/data.h b/lib/data.h
5433index 56edd34841971064ac48a0a3acae93ce5d7690aa..fcde55ada917c1c87850044c1bb79b859d6762bf 100644
5434--- a/lib/data.h
5435+++ b/lib/data.h
5436@@ -5,7 +5,8 @@
5437 /*
5438 * This word into the dictionary
5439 */
5440-typedef struct word {
5441+typedef struct word
5442+{
5443 int Id;
5444 const unsigned char *Line;
5445 } Word;
5446@@ -13,16 +14,15 @@
5447 /*
5448 * This is database connection.
5449 */
5450-typedef struct data {
5451+typedef struct data
5452+{
5453 sqlite3 *db;
5454 } Data;
5455-
5456
5457 /*
5458 * create a new data struct from sqlite filename.
5459 */
5460 Data* new_data(const char*);
5461-
5462
5463 void free_data(Data*);
5464
5465diff --git a/lib/list.c b/lib/list.c
5466index fc0fddfd3e82dc239d36563fd7953e5522aa1ff0..52feb767f49eaf205b9926a20c946d0dd570a959 100644
5467--- a/lib/list.c
5468+++ b/lib/list.c
5469@@ -4,8 +4,7 @@
5470 LIST* list_add(LIST* list, void* item)
5471 {
5472
5473- if (list == NULL)
5474- {
5475+ if (list == NULL) {
5476 list = (LIST*)malloc(sizeof(LIST));
5477 list->size = 0;
5478 list->list = (void**)malloc(sizeof(0));
5479@@ -35,20 +34,21 @@
5480 return list;
5481 }
5482
5483-void list_free(LIST* list) {
5484- for (unsigned int x = 0; x < list->size; x++)
5485+void list_free(LIST* list)
5486+{
5487+ for (unsigned int x = 0; x < list->size; x++)
5488 free(list->list[x]);
5489
5490 free(list->list);
5491 free(list);
5492 }
5493
5494-
5495-void *list_get(LIST *list, unsigned int index) {
5496+void *list_get(LIST *list, unsigned int index)
5497+{
5498 if (list == NULL)
5499 return NULL;
5500
5501- if (index < list->size)
5502+ if (index < list->size)
5503 return list->list[index];
5504
5505 return NULL;
5506diff --git a/lib/list.h b/lib/list.h
5507index dd28722a34ace46877267b9eea10edb154148bdf..18bc423838286fcccb949e25e7d164714022a852 100644
5508--- a/lib/list.h
5509+++ b/lib/list.h
5510@@ -2,7 +2,8 @@ #pragma once
5511 #include <stdlib.h>
5512
5513 #define LIST_SIZE_FACTOR 1.5
5514-struct list {
5515+struct list
5516+{
5517 unsigned int size;
5518 unsigned int allocated_size;
5519 void** list;
5520diff --git a/lib/ui.c b/lib/ui.c
5521index 1a285a00e407566b783917c1541883301ab838c7..3eae2017ce609363744cc85f4b98c797b63e35fc 100644
5522--- a/lib/ui.c
5523+++ b/lib/ui.c
5524@@ -9,7 +9,8 @@ #include "ui.h"
5525
5526 const char *uload = "█";
5527
5528-PROGRESS_BAR* new_progress_bar(WINDOW* scr, float total) {
5529+PROGRESS_BAR* new_progress_bar(WINDOW* scr, float total)
5530+{
5531 PROGRESS_BAR *bar = (PROGRESS_BAR*)malloc(sizeof(PROGRESS_BAR));
5532 bar->scr = scr;
5533 bar->total = total;
5534@@ -17,7 +18,8 @@ bar->current = 0;
5535 return bar;
5536 }
5537
5538-void bar_step(PROGRESS_BAR* bar, float step){
5539+void bar_step(PROGRESS_BAR* bar, float step)
5540+{
5541 bar->current += step;
5542
5543 int x, y;
5544@@ -42,13 +44,12 @@
5545 wmove(bar->scr, hy+1, hx - len);
5546 wprintw(bar->scr, "%.0f/%.0f", bar->current, bar->total);
5547
5548-
5549 wmove(bar->scr,0,0);
5550 wrefresh(bar->scr);
5551 }
5552
5553-
5554-TEXT_BOX* new_text_box(WINDOW* scr, int length) {
5555+TEXT_BOX* new_text_box(WINDOW* scr, int length)
5556+{
5557 TEXT_BOX *text = (TEXT_BOX*)malloc(sizeof(TEXT_BOX));
5558 text->scr = scr;
5559 text->length = length;
5560@@ -58,22 +59,23 @@ memset(text->text, '\0', length);
5561 return text;
5562 }
5563
5564-void get_char(TEXT_BOX* text, void (*sch)(char*, int)) {
5565- while(1){
5566+void get_char(TEXT_BOX* text, void (*sch)(char*, int))
5567+{
5568+ while(1) {
5569 wchar_t c;
5570 get_wch((wint_t*)&c);
5571
5572 switch(c) {
5573- case KEY_BACKSPACE:
5574- if (text->current > 0) {
5575- text->text[text->current--] = '\0';
5576- }
5577- break;
5578- default:
5579- if (text->current < (text->length-2)) {
5580- text->text[text->current] = c;
5581- text->text[++text->current] = '\0';
5582- }
5583+ case KEY_BACKSPACE:
5584+ if (text->current > 0) {
5585+ text->text[text->current--] = '\0';
5586+ }
5587+ break;
5588+ default:
5589+ if (text->current < (text->length-2)) {
5590+ text->text[text->current] = c;
5591+ text->text[++text->current] = '\0';
5592+ }
5593 }
5594
5595 char str[text->length];
5596diff --git a/lib/ui.h b/lib/ui.h
5597index 90b352ff9f0cd7dcc853c3e6695fb9ce74d2b10a..12ee2f4932a6fccf18ea2d71a7694baee3e4ea88 100644
5598--- a/lib/ui.h
5599+++ b/lib/ui.h
5600@@ -1,7 +1,8 @@
5601 #pragma once
5602 #include <ncurses.h>
5603
5604-typedef struct progress_bar {
5605+typedef struct progress_bar
5606+{
5607 float total;
5608 float current;
5609 WINDOW *scr;
5610@@ -10,8 +11,8 @@
5611 PROGRESS_BAR* new_progress_bar(WINDOW*, float);
5612 void bar_step(PROGRESS_BAR*, float);
5613
5614-
5615-typedef struct text_box {
5616+typedef struct text_box
5617+{
5618 wchar_t *text;
5619 int length;
5620 int current;
5621@@ -20,4 +21,3 @@ } TEXT_BOX;
5622
5623 TEXT_BOX* new_text_box(WINDOW*, int);
5624 void get_char(TEXT_BOX* text, void (*sch)(char*, int));
5625-
5626diff --git a/lib/util.c b/lib/util.c
5627index 6720082ac9e4f530f85d26e47416487702fac414..895ca89566c23aea5ca38733825ab012448f1f4f 100644
5628--- a/lib/util.c
5629+++ b/lib/util.c
5630@@ -6,7 +6,8 @@ #include "util.h"
5631
5632 #define BUF_SIZE 100
5633
5634-char* copy_achar(const char* src) {
5635+char* copy_achar(const char* src)
5636+{
5637 int len = strlen(src) + 1;
5638 char* dest = (char*)malloc(sizeof(char)*len);
5639 strcpy(dest, src);
5640@@ -14,8 +15,8 @@
5641 return dest;
5642 }
5643
5644-
5645-int load_or_save_db(sqlite3 *pInMemory, const char *zFilename, int isSave){
5646+int load_or_save_db(sqlite3 *pInMemory, const char *zFilename, int isSave)
5647+{
5648 int rc; /* Function return code */
5649 sqlite3 *pFile; /* Database connection opened on zFilename */
5650 sqlite3_backup *pBackup; /* Backup object used to copy data */
5651@@ -23,12 +24,12 @@ sqlite3 *pTo; /* Database to copy to (pFile or pInMemory) */
5652 sqlite3 *pFrom; /* Database to copy from (pFile or pInMemory) */
5653
5654 rc = sqlite3_open(zFilename, &pFile);
5655- if( rc==SQLITE_OK ){
5656+ if( rc==SQLITE_OK ) {
5657 pFrom = (isSave ? pInMemory : pFile);
5658 pTo = (isSave ? pFile : pInMemory);
5659
5660 pBackup = sqlite3_backup_init(pTo, "main", pFrom, "main");
5661- if( pBackup ){
5662+ if( pBackup ) {
5663 (void)sqlite3_backup_step(pBackup, -1);
5664 (void)sqlite3_backup_finish(pBackup);
5665 }
5666@@ -39,11 +40,11 @@ (void)sqlite3_close(pFile);
5667 return rc;
5668 }
5669
5670-unsigned int count_file_lines(FILE *file) {
5671+unsigned int count_file_lines(FILE *file)
5672+{
5673 char buf[BUF_SIZE];
5674 unsigned int counter = 0;
5675- for(;;)
5676- {
5677+ for(;;) {
5678 size_t res = fread(buf, 1, BUF_SIZE, file);
5679 if (ferror(file))
5680 return -1;
5681diff --git a/lib/util.h b/lib/util.h
5682index c03dbae7654b1ac748acd438ed6ab0f4975a2427..33c61ed8429051c14b8d82b102c2fbaa0d85570f 100644
5683--- a/lib/util.h
5684+++ b/lib/util.h
5685@@ -8,8 +8,6 @@ * Copy of a char to a newly created string of the same size.
5686 */
5687 char* copy_achar(const char*);
5688
5689-
5690 int load_or_save_db(sqlite3 *pInMemory, const char *zFilename, int isSave);
5691-
5692
5693 unsigned int count_file_lines(FILE *file);