dll/win32/msi/tokenize.c

   1 /*
   2 ** 2001 September 15
   3 **
   4 ** The author disclaims copyright to this source code.  In place of
   5 ** a legal notice, here is a blessing:
   6 **
   7 **    May you do good and not evil.
   8 **    May you find forgiveness for yourself and forgive others.
   9 **    May you share freely, never taking more than you give.
  10 **
  11 *************************************************************************
  12 ** A tokenizer for SQL
  13 **
  14 ** This file contains C code that splits an SQL input string up into
  15 ** individual tokens and sends those tokens one-by-one over to the
  16 ** parser for analysis.
  17 */
  18
  19 #include "msipriv.h"
  20 #include "sql.tab.h"
  21
  22 /*
  23 ** All the keywords of the SQL language are stored as in a hash
  24 ** table composed of instances of the following structure.
  25 */
  26 typedef struct Keyword Keyword;
  27 struct Keyword {
  28   const WCHAR *name;             /* The keyword name */
  29   unsigned int len;
  30   int tokenType;           /* The token value for this keyword */
  31 };
  32
  33 #define MAX_TOKEN_LEN 11
  34
  35 static const WCHAR addW[] = {'A','D','D'};
  36 static const WCHAR alterW[] = {'A','L','T','E','R'};
  37 static const WCHAR andW[] = {'A','N','D'};
  38 static const WCHAR byW[] = {'B','Y'};
  39 static const WCHAR charW[] = {'C','H','A','R'};
  40 static const WCHAR characterW[] = {'C','H','A','R','A','C','T','E','R'};
  41 static const WCHAR createW[] = {'C','R','E','A','T','E'};
  42 static const WCHAR deleteW[] = {'D','E','L','E','T','E'};
  43 static const WCHAR distinctW[] = {'D','I','S','T','I','N','C','T'};
  44 static const WCHAR dropW[] = {'D','R','O','P'};
  45 static const WCHAR freeW[] = {'F','R','E','E'};
  46 static const WCHAR fromW[] = {'F','R','O','M'};
  47 static const WCHAR holdW[] = {'H','O','L','D'};
  48 static const WCHAR insertW[] = {'I','N','S','E','R','T'};
  49 static const WCHAR intW[] = {'I','N','T'};
  50 static const WCHAR integerW[] = {'I','N','T','E','G','E','R'};
  51 static const WCHAR intoW[] = {'I','N','T','O'};
  52 static const WCHAR isW[] = {'I','S'};
  53 static const WCHAR keyW[] = {'K','E','Y'};
  54 static const WCHAR likeW[] = {'L','I','K','E'};
  55 static const WCHAR localizableW[] = {'L','O','C','A','L','I','Z','A','B','L','E'};
  56 static const WCHAR longW[] = {'L','O','N','G'};
  57 static const WCHAR longcharW[] = {'L','O','N','G','C','H','A','R'};
  58 static const WCHAR notW[] = {'N','O','T'};
  59 static const WCHAR nullW[] = {'N','U','L','L'};
  60 static const WCHAR objectW[] = {'O','B','J','E','C','T'};
  61 static const WCHAR orW[] = {'O','R'};
  62 static const WCHAR orderW[] = {'O','R','D','E','R'};
  63 static const WCHAR primaryW[] = {'P','R','I','M','A','R','Y'};
  64 static const WCHAR selectW[] = {'S','E','L','E','C','T'};
  65 static const WCHAR setW[] = {'S','E','T'};
  66 static const WCHAR shortW[] = {'S','H','O','R','T'};
  67 static const WCHAR tableW[] = {'T','A','B','L','E'};
  68 static const WCHAR temporaryW[] = {'T','E','M','P','O','R','A','R','Y'};
  69 static const WCHAR updateW[] = {'U','P','D','A','T','E'};
  70 static const WCHAR valuesW[] = {'V','A','L','U','E','S'};
  71 static const WCHAR whereW[] = {'W','H','E','R','E'};
  72
  73 /*
  74 ** These are the keywords
  75 ** They MUST be in alphabetical order
  76 */
  77 static const Keyword aKeywordTable[] = {
  78   { addW,         ARRAY_SIZE(addW),         TK_ADD },
  79   { alterW,       ARRAY_SIZE(alterW),       TK_ALTER },
  80   { andW,         ARRAY_SIZE(andW),         TK_AND },
  81   { byW,          ARRAY_SIZE(byW),          TK_BY },
  82   { charW,        ARRAY_SIZE(charW),        TK_CHAR },
  83   { characterW,   ARRAY_SIZE(characterW),   TK_CHAR },
  84   { createW,      ARRAY_SIZE(createW),      TK_CREATE },
  85   { deleteW,      ARRAY_SIZE(deleteW),      TK_DELETE },
  86   { distinctW,    ARRAY_SIZE(distinctW),    TK_DISTINCT },
  87   { dropW,        ARRAY_SIZE(dropW),        TK_DROP },
  88   { freeW,        ARRAY_SIZE(freeW),        TK_FREE },
  89   { fromW,        ARRAY_SIZE(fromW),        TK_FROM },
  90   { holdW,        ARRAY_SIZE(holdW),        TK_HOLD },
  91   { insertW,      ARRAY_SIZE(insertW),      TK_INSERT },
  92   { intW,         ARRAY_SIZE(intW),         TK_INT },
  93   { integerW,     ARRAY_SIZE(integerW),     TK_INT },
  94   { intoW,        ARRAY_SIZE(intoW),        TK_INTO },
  95   { isW,          ARRAY_SIZE(isW),          TK_IS },
  96   { keyW,         ARRAY_SIZE(keyW),         TK_KEY },
  97   { likeW,        ARRAY_SIZE(likeW),        TK_LIKE },
  98   { localizableW, ARRAY_SIZE(localizableW), TK_LOCALIZABLE },
  99   { longW,        ARRAY_SIZE(longW),        TK_LONG },
 100   { longcharW,    ARRAY_SIZE(longcharW),    TK_LONGCHAR },
 101   { notW,         ARRAY_SIZE(notW),         TK_NOT },
 102   { nullW,        ARRAY_SIZE(nullW),        TK_NULL },
 103   { objectW,      ARRAY_SIZE(objectW),      TK_OBJECT },
 104   { orW,          ARRAY_SIZE(orW),          TK_OR },
 105   { orderW,       ARRAY_SIZE(orderW),       TK_ORDER },
 106   { primaryW,     ARRAY_SIZE(primaryW),     TK_PRIMARY },
 107   { selectW,      ARRAY_SIZE(selectW),      TK_SELECT },
 108   { setW,         ARRAY_SIZE(setW),         TK_SET },
 109   { shortW,       ARRAY_SIZE(shortW),       TK_SHORT },
 110   { tableW,       ARRAY_SIZE(tableW),       TK_TABLE },
 111   { temporaryW,   ARRAY_SIZE(temporaryW),   TK_TEMPORARY },
 112   { updateW,      ARRAY_SIZE(updateW),      TK_UPDATE },
 113   { valuesW,      ARRAY_SIZE(valuesW),      TK_VALUES },
 114   { whereW,       ARRAY_SIZE(whereW),       TK_WHERE },
 115 };
 116
 117 /*
 118 ** Comparison function for binary search.
 119 */
 120 static int compKeyword(const void *m1, const void *m2){
 121   const Keyword *k1 = m1, *k2 = m2;
 122   int ret, len = min( k1->len, k2->len );
 123
 124   if ((ret = memicmpW( k1->name, k2->name, len ))) return ret;
 125   if (k1->len < k2->len) return -1;
 126   else if (k1->len > k2->len) return 1;
 127   return 0;
 128 }
 129
 130 /*
 131 ** This function looks up an identifier to determine if it is a
 132 ** keyword.  If it is a keyword, the token code of that keyword is
 133 ** returned.  If the input is not a keyword, TK_ID is returned.
 134 */
 135 static int sqliteKeywordCode(const WCHAR *z, int n){
 136   Keyword key, *r;
 137
 138   if( n>MAX_TOKEN_LEN )
 139     return TK_ID;
 140
 141   key.tokenType = 0;
 142   key.name = z;
 143   key.len = n;
 144   r = bsearch( &key, aKeywordTable, ARRAY_SIZE(aKeywordTable), sizeof(Keyword), compKeyword );
 145   if( r )
 146     return r->tokenType;
 147   return TK_ID;
 148 }
 149
 150
 151 /*
 152 ** If X is a character that can be used in an identifier then
 153 ** isIdChar[X] will be 1.  Otherwise isIdChar[X] will be 0.
 154 **
 155 ** In this implementation, an identifier can be a string of
 156 ** alphabetic characters, digits, and "_" plus any character
 157 ** with the high-order bit set.  The latter rule means that
 158 ** any sequence of UTF-8 characters or characters taken from
 159 ** an extended ISO8859 character set can form an identifier.
 160 */
 161 static const char isIdChar[] = {
 162 /* x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 xA xB xC xD xE xF */
 163     0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x */
 164     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 1x */
 165     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,  /* 2x */
 166     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0,  /* 3x */
 167     0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 4x */
 168     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1,  /* 5x */
 169     0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 6x */
 170     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,  /* 7x */
 171     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 8x */
 172     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 9x */
 173     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Ax */
 174     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Bx */
 175     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Cx */
 176     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Dx */
 177     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Ex */
 178     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Fx */
 179 };
 180
 181 /*
 182 ** WCHAR safe version of isdigit()
 183 */
 184 static inline int isDigit(WCHAR c)
 185 {
 186     return c >= '0' && c <= '9';
 187 }
 188
 189 /*
 190 ** WCHAR safe version of isspace(), except '\r'
 191 */
 192 static inline int isSpace(WCHAR c)
 193 {
 194     return c == ' ' || c == '\t' || c == '\n' || c == '\f';
 195 }
 196
 197 /*
 198 ** Return the length of the token that begins at z[0].  Return
 199 ** -1 if the token is (or might be) incomplete.  Store the token
 200 ** type in *tokenType before returning.
 201 */
 202 int sqliteGetToken(const WCHAR *z, int *tokenType, int *skip){
 203   int i;
 204
 205   *skip = 0;
 206   switch( *z ){
 207     case ' ': case '\t': case '\n': case '\f':
 208       for(i=1; isSpace(z[i]); i++){}
 209       *tokenType = TK_SPACE;
 210       return i;
 211     case '-':
 212       if( z[1]==0 ) return -1;
 213       *tokenType = TK_MINUS;
 214       return 1;
 215     case '(':
 216       *tokenType = TK_LP;
 217       return 1;
 218     case ')':
 219       *tokenType = TK_RP;
 220       return 1;
 221     case '*':
 222       *tokenType = TK_STAR;
 223       return 1;
 224     case '=':
 225       *tokenType = TK_EQ;
 226       return 1;
 227     case '<':
 228       if( z[1]=='=' ){
 229         *tokenType = TK_LE;
 230         return 2;
 231       }else if( z[1]=='>' ){
 232         *tokenType = TK_NE;
 233         return 2;
 234       }else{
 235         *tokenType = TK_LT;
 236         return 1;
 237       }
 238     case '>':
 239       if( z[1]=='=' ){
 240         *tokenType = TK_GE;
 241         return 2;
 242       }else{
 243         *tokenType = TK_GT;
 244         return 1;
 245       }
 246     case '!':
 247       if( z[1]!='=' ){
 248         *tokenType = TK_ILLEGAL;
 249         return 2;
 250       }else{
 251         *tokenType = TK_NE;
 252         return 2;
 253       }
 254     case '?':
 255       *tokenType = TK_WILDCARD;
 256       return 1;
 257     case ',':
 258       *tokenType = TK_COMMA;
 259       return 1;
 260     case '`': case '\'': {
 261       int delim = z[0];
 262       for(i=1; z[i]; i++){
 263         if( z[i]==delim )
 264           break;
 265       }
 266       if( z[i] ) i++;
 267       if( delim == '`' )
 268         *tokenType = TK_ID;
 269       else
 270         *tokenType = TK_STRING;
 271       return i;
 272     }
 273     case '.':
 274       if( !isDigit(z[1]) ){
 275         *tokenType = TK_DOT;
 276         return 1;
 277       }
 278       /* Fall through */
 279     case '0': case '1': case '2': case '3': case '4':
 280     case '5': case '6': case '7': case '8': case '9':
 281       *tokenType = TK_INTEGER;
 282       for(i=1; isDigit(z[i]); i++){}
 283       return i;
 284     case '[':
 285       for(i=1; z[i] && z[i-1]!=']'; i++){}
 286       *tokenType = TK_ID;
 287       return i;
 288     default:
 289       if( !isIdChar[*z] ){
 290         break;
 291       }
 292       for(i=1; isIdChar[z[i]]; i++){}
 293       *tokenType = sqliteKeywordCode(z, i);
 294       if( *tokenType == TK_ID && z[i] == '`' ) *skip = 1;
 295       return i;
 296   }
 297   *tokenType = TK_ILLEGAL;
 298   return 1;
 299 }