drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 1 | /* |
| 2 | ** Copyright (c) 1999, 2000 D. Richard Hipp |
| 3 | ** |
| 4 | ** This program is free software; you can redistribute it and/or |
| 5 | ** modify it under the terms of the GNU General Public |
| 6 | ** License as published by the Free Software Foundation; either |
| 7 | ** version 2 of the License, or (at your option) any later version. |
| 8 | ** |
| 9 | ** This program is distributed in the hope that it will be useful, |
| 10 | ** but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 11 | ** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 12 | ** General Public License for more details. |
| 13 | ** |
| 14 | ** You should have received a copy of the GNU General Public |
| 15 | ** License along with this library; if not, write to the |
| 16 | ** Free Software Foundation, Inc., 59 Temple Place - Suite 330, |
| 17 | ** Boston, MA 02111-1307, USA. |
| 18 | ** |
| 19 | ** Author contact information: |
| 20 | ** drh@hwaci.com |
| 21 | ** http://www.hwaci.com/drh/ |
| 22 | ** |
| 23 | ************************************************************************* |
| 24 | ** Utility functions used throughout sqlite. |
| 25 | ** |
| 26 | ** This file contains functions for allocating memory, comparing |
| 27 | ** strings, and stuff like that. |
| 28 | ** |
drh | 7c68d60 | 2000-10-11 19:28:51 +0000 | [diff] [blame^] | 29 | ** $Id: util.c,v 1.16 2000/10/11 19:28:52 drh Exp $ |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 30 | */ |
| 31 | #include "sqliteInt.h" |
| 32 | #include <stdarg.h> |
| 33 | #include <ctype.h> |
| 34 | |
drh | 7c68d60 | 2000-10-11 19:28:51 +0000 | [diff] [blame^] | 35 | /* |
| 36 | ** If MEMORY_DEBUG is defined, then use versions of malloc() and |
| 37 | ** free() that track memory usage and check for buffer overruns. |
| 38 | */ |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 39 | #ifdef MEMORY_DEBUG |
| 40 | |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 41 | /* |
| 42 | ** Allocate new memory and set it to zero. Return NULL if |
| 43 | ** no memory is available. |
| 44 | */ |
| 45 | void *sqliteMalloc_(int n, char *zFile, int line){ |
| 46 | void *p; |
| 47 | int *pi; |
| 48 | int k; |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 49 | sqlite_nMalloc++; |
| 50 | if( sqlite_iMallocFail>=0 ){ |
| 51 | sqlite_iMallocFail--; |
| 52 | if( sqlite_iMallocFail==0 ) return 0; |
| 53 | } |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 54 | k = (n+sizeof(int)-1)/sizeof(int); |
| 55 | pi = malloc( (3+k)*sizeof(int)); |
| 56 | if( pi==0 ) return 0; |
| 57 | pi[0] = 0xdead1122; |
| 58 | pi[1] = n; |
| 59 | pi[k+2] = 0xdead3344; |
| 60 | p = &pi[2]; |
| 61 | memset(p, 0, n); |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 62 | #if MEMORY_DEBUG>1 |
| 63 | fprintf(stderr,"malloc %d bytes at 0x%x from %s:%d\n", n, (int)p, zFile,line); |
| 64 | #endif |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 65 | return p; |
| 66 | } |
| 67 | |
| 68 | /* |
| 69 | ** Free memory previously obtained from sqliteMalloc() |
| 70 | */ |
| 71 | void sqliteFree_(void *p, char *zFile, int line){ |
| 72 | if( p ){ |
| 73 | int *pi, k, n; |
| 74 | pi = p; |
| 75 | pi -= 2; |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 76 | sqlite_nFree++; |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 77 | if( pi[0]!=0xdead1122 ){ |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 78 | fprintf(stderr,"Low-end memory corruption at 0x%x\n", (int)p); |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 79 | return; |
| 80 | } |
| 81 | n = pi[1]; |
| 82 | k = (n+sizeof(int)-1)/sizeof(int); |
| 83 | if( pi[k+2]!=0xdead3344 ){ |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 84 | fprintf(stderr,"High-end memory corruption at 0x%x\n", (int)p); |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 85 | return; |
| 86 | } |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 87 | memset(pi, 0xff, (k+3)*sizeof(int)); |
| 88 | #if MEMORY_DEBUG>1 |
| 89 | fprintf(stderr,"free %d bytes at 0x%x from %s:%d\n", n, (int)p, zFile,line); |
| 90 | #endif |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 91 | free(pi); |
| 92 | } |
| 93 | } |
| 94 | |
| 95 | /* |
| 96 | ** Resize a prior allocation. If p==0, then this routine |
| 97 | ** works just like sqliteMalloc(). If n==0, then this routine |
| 98 | ** works just like sqliteFree(). |
| 99 | */ |
| 100 | void *sqliteRealloc_(void *oldP, int n, char *zFile, int line){ |
| 101 | int *oldPi, *pi, k, oldN, oldK; |
| 102 | void *p; |
| 103 | if( oldP==0 ){ |
| 104 | return sqliteMalloc_(n,zFile,line); |
| 105 | } |
| 106 | if( n==0 ){ |
| 107 | sqliteFree_(oldP,zFile,line); |
| 108 | return 0; |
| 109 | } |
| 110 | oldPi = oldP; |
| 111 | oldPi -= 2; |
| 112 | if( oldPi[0]!=0xdead1122 ){ |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 113 | fprintf(stderr,"Low-end memory corruption in realloc at 0x%x\n", (int)p); |
drh | 9bb61fe | 2000-06-05 16:01:39 +0000 | [diff] [blame] | 114 | return 0; |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 115 | } |
| 116 | oldN = oldPi[1]; |
| 117 | oldK = (oldN+sizeof(int)-1)/sizeof(int); |
| 118 | if( oldPi[oldK+2]!=0xdead3344 ){ |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 119 | fprintf(stderr,"High-end memory corruption in realloc at 0x%x\n", (int)p); |
drh | 9bb61fe | 2000-06-05 16:01:39 +0000 | [diff] [blame] | 120 | return 0; |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 121 | } |
| 122 | k = (n + sizeof(int) - 1)/sizeof(int); |
| 123 | pi = malloc( (k+3)*sizeof(int) ); |
| 124 | pi[0] = 0xdead1122; |
| 125 | pi[1] = n; |
| 126 | pi[k+2] = 0xdead3344; |
| 127 | p = &pi[2]; |
| 128 | memcpy(p, oldP, n>oldN ? oldN : n); |
| 129 | if( n>oldN ){ |
| 130 | memset(&((char*)p)[oldN], 0, n-oldN); |
| 131 | } |
| 132 | memset(oldPi, 0, (oldK+3)*sizeof(int)); |
| 133 | free(oldPi); |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 134 | #if MEMORY_DEBUG>1 |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 135 | fprintf(stderr,"realloc %d to %d bytes at 0x%x to 0x%x at %s:%d\n", oldN, n, |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 136 | (int)oldP, (int)p, zFile, line); |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 137 | #endif |
drh | dcc581c | 2000-05-30 13:44:19 +0000 | [diff] [blame] | 138 | return p; |
| 139 | } |
drh | c3c2fc9 | 2000-05-31 22:58:39 +0000 | [diff] [blame] | 140 | |
| 141 | /* |
| 142 | ** Make a duplicate of a string into memory obtained from malloc() |
| 143 | ** Free the original string using sqliteFree(). |
| 144 | */ |
| 145 | void sqliteStrRealloc(char **pz){ |
| 146 | char *zNew; |
| 147 | if( pz==0 || *pz==0 ) return; |
| 148 | zNew = malloc( strlen(*pz) + 1 ); |
| 149 | if( zNew ) strcpy(zNew, *pz); |
| 150 | sqliteFree(*pz); |
| 151 | *pz = zNew; |
| 152 | } |
| 153 | |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 154 | /* |
| 155 | ** Make a copy of a string in memory obtained from sqliteMalloc() |
| 156 | */ |
| 157 | char *sqliteStrDup_(const char *z, char *zFile, int line){ |
| 158 | char *zNew = sqliteMalloc_(strlen(z)+1, zFile, line); |
| 159 | if( zNew ) strcpy(zNew, z); |
| 160 | return zNew; |
| 161 | } |
| 162 | char *sqliteStrNDup_(const char *z, int n, char *zFile, int line){ |
| 163 | char *zNew = sqliteMalloc_(n+1, zFile, line); |
| 164 | if( zNew ){ |
| 165 | memcpy(zNew, z, n); |
| 166 | zNew[n] = 0; |
| 167 | } |
| 168 | return zNew; |
| 169 | } |
drh | 7c68d60 | 2000-10-11 19:28:51 +0000 | [diff] [blame^] | 170 | #endif /* MEMORY_DEBUG */ |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 171 | |
drh | 7c68d60 | 2000-10-11 19:28:51 +0000 | [diff] [blame^] | 172 | /* |
| 173 | ** The following versions of malloc() and free() are for use in a |
| 174 | ** normal build. |
| 175 | */ |
| 176 | #if !defined(MEMORY_DEBUG) |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 177 | |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 178 | /* |
| 179 | ** Allocate new memory and set it to zero. Return NULL if |
| 180 | ** no memory is available. |
| 181 | */ |
| 182 | void *sqliteMalloc(int n){ |
| 183 | void *p = malloc(n); |
| 184 | if( p==0 ) return 0; |
| 185 | memset(p, 0, n); |
| 186 | return p; |
| 187 | } |
| 188 | |
| 189 | /* |
| 190 | ** Free memory previously obtained from sqliteMalloc() |
| 191 | */ |
| 192 | void sqliteFree(void *p){ |
drh | 305cea6 | 2000-05-29 17:44:25 +0000 | [diff] [blame] | 193 | if( p ){ |
drh | 305cea6 | 2000-05-29 17:44:25 +0000 | [diff] [blame] | 194 | free(p); |
| 195 | } |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 196 | } |
| 197 | |
| 198 | /* |
| 199 | ** Resize a prior allocation. If p==0, then this routine |
| 200 | ** works just like sqliteMalloc(). If n==0, then this routine |
| 201 | ** works just like sqliteFree(). |
| 202 | */ |
| 203 | void *sqliteRealloc(void *p, int n){ |
| 204 | if( p==0 ){ |
| 205 | return sqliteMalloc(n); |
| 206 | } |
| 207 | if( n==0 ){ |
| 208 | sqliteFree(p); |
| 209 | return 0; |
| 210 | } |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 211 | return realloc(p, n); |
| 212 | } |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 213 | |
| 214 | /* |
| 215 | ** Make a copy of a string in memory obtained from sqliteMalloc() |
| 216 | */ |
| 217 | char *sqliteStrDup(const char *z){ |
| 218 | char *zNew = sqliteMalloc(strlen(z)+1); |
| 219 | if( zNew ) strcpy(zNew, z); |
| 220 | return zNew; |
| 221 | } |
| 222 | char *sqliteStrNDup(const char *z, int n){ |
| 223 | char *zNew = sqliteMalloc(n+1); |
| 224 | if( zNew ){ |
| 225 | memcpy(zNew, z, n); |
| 226 | zNew[n] = 0; |
| 227 | } |
| 228 | return zNew; |
| 229 | } |
drh | 7c68d60 | 2000-10-11 19:28:51 +0000 | [diff] [blame^] | 230 | #endif /* !defined(MEMORY_DEBUG) */ |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 231 | |
| 232 | /* |
| 233 | ** Create a string from the 2nd and subsequent arguments (up to the |
| 234 | ** first NULL argument), store the string in memory obtained from |
| 235 | ** sqliteMalloc() and make the pointer indicated by the 1st argument |
| 236 | ** point to that string. |
| 237 | */ |
| 238 | void sqliteSetString(char **pz, const char *zFirst, ...){ |
| 239 | va_list ap; |
| 240 | int nByte; |
| 241 | const char *z; |
| 242 | char *zResult; |
| 243 | |
| 244 | if( pz==0 ) return; |
| 245 | nByte = strlen(zFirst) + 1; |
| 246 | va_start(ap, zFirst); |
| 247 | while( (z = va_arg(ap, const char*))!=0 ){ |
| 248 | nByte += strlen(z); |
| 249 | } |
| 250 | va_end(ap); |
| 251 | sqliteFree(*pz); |
| 252 | *pz = zResult = sqliteMalloc( nByte ); |
| 253 | if( zResult==0 ) return; |
| 254 | strcpy(zResult, zFirst); |
| 255 | zResult += strlen(zResult); |
| 256 | va_start(ap, zFirst); |
| 257 | while( (z = va_arg(ap, const char*))!=0 ){ |
| 258 | strcpy(zResult, z); |
| 259 | zResult += strlen(zResult); |
| 260 | } |
| 261 | va_end(ap); |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 262 | #ifdef MEMORY_DEBUG |
| 263 | #if MEMORY_DEBUG>1 |
| 264 | fprintf(stderr,"string at 0x%x is %s\n", (int)*pz, *pz); |
| 265 | #endif |
| 266 | #endif |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 267 | } |
| 268 | |
| 269 | /* |
| 270 | ** Works like sqliteSetString, but each string is now followed by |
| 271 | ** a length integer. -1 means use the whole string. |
| 272 | */ |
| 273 | void sqliteSetNString(char **pz, ...){ |
| 274 | va_list ap; |
| 275 | int nByte; |
| 276 | const char *z; |
| 277 | char *zResult; |
| 278 | int n; |
| 279 | |
| 280 | if( pz==0 ) return; |
| 281 | nByte = 0; |
| 282 | va_start(ap, pz); |
| 283 | while( (z = va_arg(ap, const char*))!=0 ){ |
| 284 | n = va_arg(ap, int); |
| 285 | if( n<=0 ) n = strlen(z); |
| 286 | nByte += n; |
| 287 | } |
| 288 | va_end(ap); |
| 289 | sqliteFree(*pz); |
| 290 | *pz = zResult = sqliteMalloc( nByte + 1 ); |
| 291 | if( zResult==0 ) return; |
| 292 | va_start(ap, pz); |
| 293 | while( (z = va_arg(ap, const char*))!=0 ){ |
| 294 | n = va_arg(ap, int); |
| 295 | if( n<=0 ) n = strlen(z); |
| 296 | strncpy(zResult, z, n); |
| 297 | zResult += n; |
| 298 | } |
| 299 | *zResult = 0; |
drh | 6e142f5 | 2000-06-08 13:36:40 +0000 | [diff] [blame] | 300 | #ifdef MEMORY_DEBUG |
| 301 | #if MEMORY_DEBUG>1 |
| 302 | fprintf(stderr,"string at 0x%x is %s\n", (int)*pz, *pz); |
| 303 | #endif |
| 304 | #endif |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 305 | va_end(ap); |
| 306 | } |
| 307 | |
drh | 982cef7 | 2000-05-30 16:27:03 +0000 | [diff] [blame] | 308 | /* |
| 309 | ** Convert an SQL-style quoted string into a normal string by removing |
| 310 | ** the quote characters. The conversion is done in-place. If the |
| 311 | ** input does not begin with a quote character, then this routine |
| 312 | ** is a no-op. |
| 313 | */ |
| 314 | void sqliteDequote(char *z){ |
| 315 | int quote; |
| 316 | int i, j; |
| 317 | quote = z[0]; |
| 318 | if( quote!='\'' && quote!='"' ) return; |
| 319 | for(i=1, j=0; z[i]; i++){ |
| 320 | if( z[i]==quote ){ |
| 321 | if( z[i+1]==quote ){ |
| 322 | z[j++] = quote; |
| 323 | i++; |
| 324 | }else{ |
| 325 | z[j++] = 0; |
| 326 | break; |
| 327 | } |
| 328 | }else{ |
| 329 | z[j++] = z[i]; |
| 330 | } |
| 331 | } |
| 332 | } |
| 333 | |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 334 | /* An array to map all upper-case characters into their corresponding |
| 335 | ** lower-case character. |
| 336 | */ |
| 337 | static unsigned char UpperToLower[] = { |
| 338 | 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, |
| 339 | 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, |
| 340 | 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, |
| 341 | 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 97, 98, 99,100,101,102,103, |
| 342 | 104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121, |
| 343 | 122, 91, 92, 93, 94, 95, 96, 97, 98, 99,100,101,102,103,104,105,106,107, |
| 344 | 108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125, |
| 345 | 126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143, |
| 346 | 144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161, |
| 347 | 162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179, |
| 348 | 180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197, |
| 349 | 198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215, |
| 350 | 216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233, |
| 351 | 234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251, |
| 352 | 252,253,254,255 |
| 353 | }; |
| 354 | |
| 355 | /* |
| 356 | ** This function computes a hash on the name of a keyword. |
| 357 | ** Case is not significant. |
| 358 | */ |
| 359 | int sqliteHashNoCase(const char *z, int n){ |
| 360 | int h = 0; |
| 361 | int c; |
| 362 | if( n<=0 ) n = strlen(z); |
| 363 | while( n-- > 0 && (c = *z++)!=0 ){ |
| 364 | h = h<<3 ^ h ^ UpperToLower[c]; |
| 365 | } |
| 366 | if( h<0 ) h = -h; |
| 367 | return h; |
| 368 | } |
| 369 | |
| 370 | /* |
drh | 967e8b7 | 2000-06-21 13:59:10 +0000 | [diff] [blame] | 371 | ** Some systems have stricmp(). Others have strcasecmp(). Because |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 372 | ** there is no consistency, we will define our own. |
| 373 | */ |
| 374 | int sqliteStrICmp(const char *zLeft, const char *zRight){ |
| 375 | register unsigned char *a, *b; |
| 376 | a = (unsigned char *)zLeft; |
| 377 | b = (unsigned char *)zRight; |
| 378 | while( *a!=0 && UpperToLower[*a]==UpperToLower[*b]){ a++; b++; } |
| 379 | return *a - *b; |
| 380 | } |
| 381 | int sqliteStrNICmp(const char *zLeft, const char *zRight, int N){ |
| 382 | register unsigned char *a, *b; |
| 383 | a = (unsigned char *)zLeft; |
| 384 | b = (unsigned char *)zRight; |
| 385 | while( N-- > 0 && *a!=0 && UpperToLower[*a]==UpperToLower[*b]){ a++; b++; } |
drh | bec2bf4 | 2000-05-29 23:48:22 +0000 | [diff] [blame] | 386 | return N<0 ? 0 : *a - *b; |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 387 | } |
| 388 | |
| 389 | /* Notes on string comparisions. |
| 390 | ** |
| 391 | ** We want the main string comparision function used for sorting to |
| 392 | ** sort both numbers and alphanumeric words into the correct sequence. |
| 393 | ** The same routine should do both without prior knowledge of which |
| 394 | ** type of text the input represents. It should even work for strings |
| 395 | ** which are a mixture of text and numbers. |
| 396 | ** |
| 397 | ** To accomplish this, we keep track of a state number while scanning |
| 398 | ** the two strings. The states are as follows: |
| 399 | ** |
| 400 | ** 1 Beginning of word |
| 401 | ** 2 Arbitrary text |
| 402 | ** 3 Integer |
| 403 | ** 4 Negative integer |
| 404 | ** 5 Real number |
| 405 | ** 6 Negative real |
| 406 | ** |
| 407 | ** The scan begins in state 1, beginning of word. Transitions to other |
| 408 | ** states are determined by characters seen, as shown in the following |
| 409 | ** chart: |
| 410 | ** |
| 411 | ** Current State Character Seen New State |
| 412 | ** -------------------- -------------- ------------------- |
| 413 | ** 0 Beginning of word "-" 3 Negative integer |
| 414 | ** digit 2 Integer |
| 415 | ** space 0 Beginning of word |
| 416 | ** otherwise 1 Arbitrary text |
| 417 | ** |
| 418 | ** 1 Arbitrary text space 0 Beginning of word |
| 419 | ** digit 2 Integer |
| 420 | ** otherwise 1 Arbitrary text |
| 421 | ** |
| 422 | ** 2 Integer space 0 Beginning of word |
| 423 | ** "." 4 Real number |
| 424 | ** digit 2 Integer |
| 425 | ** otherwise 1 Arbitrary text |
| 426 | ** |
| 427 | ** 3 Negative integer space 0 Beginning of word |
| 428 | ** "." 5 Negative Real num |
| 429 | ** digit 3 Negative integer |
| 430 | ** otherwise 1 Arbitrary text |
| 431 | ** |
| 432 | ** 4 Real number space 0 Beginning of word |
| 433 | ** digit 4 Real number |
| 434 | ** otherwise 1 Arbitrary text |
| 435 | ** |
| 436 | ** 5 Negative real num space 0 Beginning of word |
| 437 | ** digit 5 Negative real num |
| 438 | ** otherwise 1 Arbitrary text |
| 439 | ** |
| 440 | ** To implement this state machine, we first classify each character |
| 441 | ** into on of the following categories: |
| 442 | ** |
| 443 | ** 0 Text |
| 444 | ** 1 Space |
| 445 | ** 2 Digit |
| 446 | ** 3 "-" |
| 447 | ** 4 "." |
| 448 | ** |
| 449 | ** Given an arbitrary character, the array charClass[] maps that character |
| 450 | ** into one of the atove categories. |
| 451 | */ |
| 452 | static const unsigned char charClass[] = { |
| 453 | /* x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 xA xB xC xD xE xF */ |
| 454 | /* 0x */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0, |
| 455 | /* 1x */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 456 | /* 2x */ 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 4, 0, |
| 457 | /* 3x */ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 0, 0, 0, 0, 0, 0, |
| 458 | /* 4x */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 459 | /* 5x */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 460 | /* 6x */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 461 | /* 7x */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 462 | /* 8x */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 463 | /* 9x */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 464 | /* Ax */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 465 | /* Bx */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 466 | /* Cx */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 467 | /* Dx */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 468 | /* Ex */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 469 | /* Fx */ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 470 | }; |
| 471 | #define N_CHAR_CLASS 5 |
| 472 | |
| 473 | /* |
| 474 | ** Given the current state number (0 thru 5), this array figures |
| 475 | ** the new state number given the character class. |
| 476 | */ |
| 477 | static const unsigned char stateMachine[] = { |
| 478 | /* Text, Space, Digit, "-", "." */ |
| 479 | 1, 0, 2, 3, 1, /* State 0: Beginning of word */ |
| 480 | 1, 0, 2, 1, 1, /* State 1: Arbitrary text */ |
| 481 | 1, 0, 2, 1, 4, /* State 2: Integer */ |
| 482 | 1, 0, 3, 1, 5, /* State 3: Negative integer */ |
| 483 | 1, 0, 4, 1, 1, /* State 4: Real number */ |
| 484 | 1, 0, 5, 1, 1, /* State 5: Negative real num */ |
| 485 | }; |
| 486 | |
| 487 | /* This routine does a comparison of two strings. Case is used only |
| 488 | ** if useCase!=0. Numbers compare in numerical order. |
| 489 | */ |
| 490 | static int privateStrCmp(const char *atext, const char *btext, int useCase){ |
| 491 | register unsigned char *a, *b, *map, ca, cb; |
| 492 | int result; |
| 493 | register int cclass = 0; |
| 494 | |
| 495 | a = (unsigned char *)atext; |
| 496 | b = (unsigned char *)btext; |
| 497 | if( useCase ){ |
| 498 | do{ |
| 499 | if( (ca= *a++)!=(cb= *b++) ) break; |
| 500 | cclass = stateMachine[cclass*N_CHAR_CLASS + charClass[ca]]; |
| 501 | }while( ca!=0 ); |
| 502 | }else{ |
| 503 | map = UpperToLower; |
| 504 | do{ |
| 505 | if( (ca=map[*a++])!=(cb=map[*b++]) ) break; |
| 506 | cclass = stateMachine[cclass*N_CHAR_CLASS + charClass[ca]]; |
| 507 | }while( ca!=0 ); |
| 508 | } |
| 509 | switch( cclass ){ |
| 510 | case 0: |
| 511 | case 1: { |
| 512 | if( isdigit(ca) && isdigit(cb) ){ |
| 513 | cclass = 2; |
| 514 | } |
| 515 | break; |
| 516 | } |
| 517 | default: { |
| 518 | break; |
| 519 | } |
| 520 | } |
| 521 | switch( cclass ){ |
| 522 | case 2: |
| 523 | case 3: { |
| 524 | if( isdigit(ca) ){ |
| 525 | if( isdigit(cb) ){ |
| 526 | int acnt, bcnt; |
| 527 | acnt = bcnt = 0; |
| 528 | while( isdigit(*a++) ) acnt++; |
| 529 | while( isdigit(*b++) ) bcnt++; |
| 530 | result = acnt - bcnt; |
| 531 | if( result==0 ) result = ca-cb; |
| 532 | }else{ |
| 533 | result = 1; |
| 534 | } |
| 535 | }else if( isdigit(cb) ){ |
| 536 | result = -1; |
| 537 | }else if( ca=='.' ){ |
| 538 | result = 1; |
| 539 | }else if( cb=='.' ){ |
| 540 | result = -1; |
| 541 | }else{ |
| 542 | result = ca - cb; |
| 543 | cclass = 2; |
| 544 | } |
| 545 | if( cclass==3 ) result = -result; |
| 546 | break; |
| 547 | } |
| 548 | case 0: |
| 549 | case 1: |
| 550 | case 4: { |
| 551 | result = ca - cb; |
| 552 | break; |
| 553 | } |
| 554 | case 5: { |
| 555 | result = cb - ca; |
| 556 | }; |
| 557 | } |
| 558 | return result; |
| 559 | } |
| 560 | |
drh | a5c2ad0 | 2000-09-14 01:21:10 +0000 | [diff] [blame] | 561 | /* |
| 562 | ** Do a comparison of pure numerics. If either string is not a pure |
| 563 | ** numeric, then return 0. Otherwise return 1 and set *pResult to be |
| 564 | ** negative, zero or positive if the first string are numerially less than |
| 565 | ** equal to, or greater than the second. |
| 566 | */ |
| 567 | static int privateCompareNum(const char *a, const char *b, int *pResult){ |
| 568 | char *endPtr; |
| 569 | double rA, rB; |
| 570 | int isNumA, isNumB; |
| 571 | if( isdigit(*a) || ((*a=='-' || *a=='+') && isdigit(a[1])) ){ |
| 572 | rA = strtod(a, &endPtr); |
| 573 | isNumA = *endPtr==0; |
| 574 | }else{ |
| 575 | isNumA = 0; |
| 576 | } |
| 577 | if( isdigit(*b) || ((*b=='-' || *b=='+') && isdigit(b[1])) ){ |
| 578 | rB = strtod(b, &endPtr); |
| 579 | isNumB = *endPtr==0; |
| 580 | }else{ |
| 581 | isNumB = 0; |
| 582 | } |
| 583 | if( isNumB==0 && isNumA==0 ) return 0; |
| 584 | if( isNumA!=isNumB ){ |
| 585 | *pResult = isNumA - isNumB; |
| 586 | }else if( rA<rB ){ |
| 587 | *pResult = -1; |
| 588 | }else if( rA>rB ){ |
| 589 | *pResult = 1; |
| 590 | }else{ |
| 591 | *pResult = 0; |
| 592 | } |
| 593 | return 1; |
| 594 | } |
| 595 | |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 596 | /* This comparison routine is what we use for comparison operations |
| 597 | ** in an SQL expression. (Ex: name<'Hello' or value<5). Compare two |
| 598 | ** strings. Use case only as a tie-breaker. Numbers compare in |
| 599 | ** numerical order. |
| 600 | */ |
| 601 | int sqliteCompare(const char *atext, const char *btext){ |
| 602 | int result; |
drh | a5c2ad0 | 2000-09-14 01:21:10 +0000 | [diff] [blame] | 603 | if( !privateCompareNum(atext, btext, &result) || result==0 ){ |
| 604 | result = privateStrCmp(atext, btext, 0); |
| 605 | if( result==0 ) result = privateStrCmp(atext, btext, 1); |
| 606 | } |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 607 | return result; |
| 608 | } |
| 609 | |
| 610 | /* |
| 611 | ** If you compile just this one file with the -DTEST_COMPARE=1 option, |
| 612 | ** it generates a program to test the comparisons routines. |
| 613 | */ |
| 614 | #ifdef TEST_COMPARE |
| 615 | #include <stdlib.h> |
| 616 | #include <stdio.h> |
| 617 | int sortCmp(const char **a, const char **b){ |
| 618 | return sqliteCompare(*a, *b); |
| 619 | } |
| 620 | int main(int argc, char **argv){ |
drh | a5c2ad0 | 2000-09-14 01:21:10 +0000 | [diff] [blame] | 621 | int i, j, k, n, cnt; |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 622 | static char *azStr[] = { |
| 623 | "abc", "aBc", "abcd", "aBcd", |
drh | a5c2ad0 | 2000-09-14 01:21:10 +0000 | [diff] [blame] | 624 | "123", "124", "1234", "-123", "-124", "-1234", "+124", |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 625 | "123.45", "123.456", "123.46", "-123.45", "-123.46", "-123.456", |
| 626 | "x9", "x10", "x-9", "x-10", "X9", "X10", |
drh | a5c2ad0 | 2000-09-14 01:21:10 +0000 | [diff] [blame] | 627 | "1.234e+02", "+123", "1.23E2", "1.2345e+2", "-1.2345e2", "+w" |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 628 | }; |
| 629 | n = sizeof(azStr)/sizeof(azStr[0]); |
| 630 | qsort(azStr, n, sizeof(azStr[0]), sortCmp); |
| 631 | for(i=0; i<n; i++){ |
| 632 | printf("%s\n", azStr[i]); |
| 633 | } |
| 634 | printf("Sanity1..."); |
| 635 | fflush(stdout); |
drh | a5c2ad0 | 2000-09-14 01:21:10 +0000 | [diff] [blame] | 636 | cnt = 0; |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 637 | for(i=0; i<n-1; i++){ |
| 638 | char *a = azStr[i]; |
| 639 | for(j=i+1; j<n; j++){ |
| 640 | char *b = azStr[j]; |
| 641 | if( sqliteCompare(a,b) != -sqliteCompare(b,a) ){ |
| 642 | printf("Failed! \"%s\" vs \"%s\"\n", a, b); |
| 643 | i = j = n; |
| 644 | } |
drh | a5c2ad0 | 2000-09-14 01:21:10 +0000 | [diff] [blame] | 645 | cnt++; |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 646 | } |
| 647 | } |
| 648 | if( i<n ){ |
drh | a5c2ad0 | 2000-09-14 01:21:10 +0000 | [diff] [blame] | 649 | printf(" OK (%d)\n", cnt); |
| 650 | } |
| 651 | printf("Sanity2..."); |
| 652 | fflush(stdout); |
| 653 | cnt = 0; |
| 654 | for(i=0; i<n; i++){ |
| 655 | char *a = azStr[i]; |
| 656 | for(j=0; j<n; j++){ |
| 657 | char *b = azStr[j]; |
| 658 | for(k=0; k<n; k++){ |
| 659 | char *c = azStr[k]; |
| 660 | int x1, x2, x3, success; |
| 661 | x1 = sqliteCompare(a,b); |
| 662 | x2 = sqliteCompare(b,c); |
| 663 | x3 = sqliteCompare(a,c); |
| 664 | if( x1==0 ){ |
| 665 | success = x2==x3; |
| 666 | }else if( x1<0 ){ |
| 667 | success = (x2<=0 && x3<=0) || x2>0; |
| 668 | }else{ |
| 669 | success = (x2>=0 && x3>=0) || x2<0; |
| 670 | } |
| 671 | if( !success ){ |
| 672 | printf("Failed! \"%s\" vs \"%s\" vs \"%s\"\n", a, b, c); |
| 673 | i = j = k = n+1; |
| 674 | } |
| 675 | cnt++; |
| 676 | } |
| 677 | } |
| 678 | } |
| 679 | if( i<n+1 ){ |
| 680 | printf(" OK (%d)\n", cnt); |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 681 | } |
| 682 | return 0; |
| 683 | } |
| 684 | #endif |
| 685 | |
| 686 | /* |
drh | 16e5955 | 2000-07-31 11:57:37 +0000 | [diff] [blame] | 687 | ** This routine is used for sorting. Each key is a list of one or more |
| 688 | ** null-terminated strings. The list is terminated by two nulls in |
| 689 | ** a row. For example, the following text is key with three strings: |
drh | 7589723 | 2000-05-29 14:26:00 +0000 | [diff] [blame] | 690 | ** |
| 691 | ** +one\000-two\000+three\000\000 |
| 692 | ** |
| 693 | ** Both arguments will have the same number of strings. This routine |
| 694 | ** returns negative, zero, or positive if the first argument is less |
| 695 | ** than, equal to, or greater than the first. (Result is a-b). |
| 696 | ** |
| 697 | ** Every string begins with either a "+" or "-" character. If the |
| 698 | ** character is "-" then the return value is negated. This is done |
| 699 | ** to implement a sort in descending order. |
| 700 | */ |
| 701 | int sqliteSortCompare(const char *a, const char *b){ |
| 702 | int len; |
| 703 | int res = 0; |
| 704 | |
| 705 | while( res==0 && *a && *b ){ |
| 706 | res = sqliteCompare(&a[1], &b[1]); |
| 707 | if( res==0 ){ |
| 708 | len = strlen(a) + 1; |
| 709 | a += len; |
| 710 | b += len; |
| 711 | } |
| 712 | } |
| 713 | if( *a=='-' ) res = -res; |
| 714 | return res; |
| 715 | } |
drh | dce2cbe | 2000-05-31 02:27:49 +0000 | [diff] [blame] | 716 | |
| 717 | /* |
| 718 | ** Compare two strings for equality where the first string can |
| 719 | ** potentially be a "glob" expression. Return true (1) if they |
| 720 | ** are the same and false (0) if they are different. |
| 721 | ** |
| 722 | ** Globbing rules: |
| 723 | ** |
| 724 | ** '*' Matches any sequence of zero or more characters. |
| 725 | ** |
| 726 | ** '?' Matches exactly one character. |
| 727 | ** |
| 728 | ** [...] Matches one character from the enclosed list of |
| 729 | ** characters. |
| 730 | ** |
| 731 | ** [^...] Matches one character not in the enclosed list. |
| 732 | ** |
| 733 | ** With the [...] and [^...] matching, a ']' character can be included |
| 734 | ** in the list by making it the first character after '[' or '^'. A |
| 735 | ** range of characters can be specified using '-'. Example: |
| 736 | ** "[a-z]" matches any single lower-case letter. To match a '-', make |
| 737 | ** it the last character in the list. |
| 738 | ** |
| 739 | ** This routine is usually quick, but can be N**2 in the worst case. |
| 740 | ** |
| 741 | ** Hints: to match '*' or '?', put them in "[]". Like this: |
| 742 | ** |
| 743 | ** abc[*]xyz Matches "abc*xyz" only |
| 744 | */ |
| 745 | int sqliteGlobCompare(const char *zPattern, const char *zString){ |
| 746 | register char c; |
| 747 | int invert; |
| 748 | int seen; |
| 749 | char c2; |
| 750 | |
| 751 | while( (c = *zPattern)!=0 ){ |
| 752 | switch( c ){ |
| 753 | case '*': |
| 754 | while( zPattern[1]=='*' ) zPattern++; |
| 755 | if( zPattern[1]==0 ) return 1; |
| 756 | c = zPattern[1]; |
| 757 | if( c=='[' || c=='?' ){ |
| 758 | while( *zString && sqliteGlobCompare(&zPattern[1],zString)==0 ){ |
| 759 | zString++; |
| 760 | } |
| 761 | return *zString!=0; |
| 762 | }else{ |
| 763 | while( (c2 = *zString)!=0 ){ |
| 764 | while( c2 != 0 && c2 != c ){ c2 = *++zString; } |
drh | c61053b | 2000-06-04 12:58:36 +0000 | [diff] [blame] | 765 | if( c2==0 ) return 0; |
drh | dce2cbe | 2000-05-31 02:27:49 +0000 | [diff] [blame] | 766 | if( sqliteGlobCompare(&zPattern[1],zString) ) return 1; |
| 767 | zString++; |
| 768 | } |
| 769 | return 0; |
| 770 | } |
| 771 | case '?': |
| 772 | if( *zString==0 ) return 0; |
| 773 | break; |
| 774 | case '[': |
| 775 | seen = 0; |
| 776 | invert = 0; |
| 777 | c = *zString; |
| 778 | if( c==0 ) return 0; |
| 779 | c2 = *++zPattern; |
| 780 | if( c2=='^' ){ invert = 1; c2 = *++zPattern; } |
| 781 | if( c2==']' ){ |
| 782 | if( c==']' ) seen = 1; |
| 783 | c2 = *++zPattern; |
| 784 | } |
| 785 | while( (c2 = *zPattern)!=0 && c2!=']' ){ |
| 786 | if( c2=='-' && zPattern[1]!=']' && zPattern[1]!=0 ){ |
| 787 | if( c>zPattern[-1] && c<zPattern[1] ) seen = 1; |
| 788 | }else if( c==c2 ){ |
| 789 | seen = 1; |
| 790 | } |
| 791 | zPattern++; |
| 792 | } |
| 793 | if( c2==0 || (seen ^ invert)==0 ) return 0; |
| 794 | break; |
| 795 | default: |
| 796 | if( c != *zString ) return 0; |
| 797 | break; |
| 798 | } |
| 799 | zPattern++; |
| 800 | zString++; |
| 801 | } |
| 802 | return *zString==0; |
| 803 | } |
| 804 | |
| 805 | /* |
| 806 | ** Compare two strings for equality using the "LIKE" operator of |
| 807 | ** SQL. The '%' character matches any sequence of 0 or more |
| 808 | ** characters and '_' matches any single character. Case is |
| 809 | ** not significant. |
| 810 | ** |
| 811 | ** This routine is just an adaptation of the sqliteGlobCompare() |
| 812 | ** routine above. |
| 813 | */ |
| 814 | int |
| 815 | sqliteLikeCompare(const unsigned char *zPattern, const unsigned char *zString){ |
| 816 | register char c; |
drh | dce2cbe | 2000-05-31 02:27:49 +0000 | [diff] [blame] | 817 | char c2; |
| 818 | |
| 819 | while( (c = UpperToLower[*zPattern])!=0 ){ |
| 820 | switch( c ){ |
| 821 | case '%': |
| 822 | while( zPattern[1]=='%' ) zPattern++; |
| 823 | if( zPattern[1]==0 ) return 1; |
| 824 | c = UpperToLower[0xff & zPattern[1]]; |
| 825 | if( c=='_' ){ |
| 826 | while( *zString && sqliteLikeCompare(&zPattern[1],zString)==0 ){ |
| 827 | zString++; |
| 828 | } |
| 829 | return *zString!=0; |
| 830 | }else{ |
| 831 | while( (c2 = UpperToLower[*zString])!=0 ){ |
| 832 | while( c2 != 0 && c2 != c ){ c2 = UpperToLower[*++zString]; } |
drh | c61053b | 2000-06-04 12:58:36 +0000 | [diff] [blame] | 833 | if( c2==0 ) return 0; |
drh | dce2cbe | 2000-05-31 02:27:49 +0000 | [diff] [blame] | 834 | if( sqliteLikeCompare(&zPattern[1],zString) ) return 1; |
| 835 | zString++; |
| 836 | } |
| 837 | return 0; |
| 838 | } |
| 839 | case '_': |
| 840 | if( *zString==0 ) return 0; |
| 841 | break; |
| 842 | default: |
| 843 | if( c != UpperToLower[*zString] ) return 0; |
| 844 | break; |
| 845 | } |
| 846 | zPattern++; |
| 847 | zString++; |
| 848 | } |
| 849 | return *zString==0; |
| 850 | } |