dlls/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 <ctype.h>
  20 #include <stdarg.h>
  21 #include <stdlib.h>
  22
  23 #include "windef.h"
  24 #include "winbase.h"
  25 #include "wine/unicode.h"
  26 #include "query.h"
  27 #include "sql.tab.h"
  28
  29 /*
  30 ** All the keywords of the SQL language are stored as in a hash
  31 ** table composed of instances of the following structure.
  32 */
  33 typedef struct Keyword Keyword;
  34 struct Keyword {
  35   const WCHAR *zName;             /* The keyword name */
  36   int tokenType;           /* The token value for this keyword */
  37 };
  38
  39 #define MAX_TOKEN_LEN 11
  40
  41 static const WCHAR ADD_W[] = { 'A','D','D',0 };
  42 static const WCHAR ALTER_W[] = { 'A','L','T','E','R',0 };
  43 static const WCHAR AND_W[] = { 'A','N','D',0 };
  44 static const WCHAR BY_W[] = { 'B','Y',0 };
  45 static const WCHAR CHAR_W[] = { 'C','H','A','R',0 };
  46 static const WCHAR CHARACTER_W[] = { 'C','H','A','R','A','C','T','E','R',0 };
  47 static const WCHAR CREATE_W[] = { 'C','R','E','A','T','E',0 };
  48 static const WCHAR DELETE_W[] = { 'D','E','L','E','T','E',0 };
  49 static const WCHAR DISTINCT_W[] = { 'D','I','S','T','I','N','C','T',0 };
  50 static const WCHAR DROP_W[] = { 'D','R','O','P',0 };
  51 static const WCHAR FREE_W[] = { 'F','R','E','E',0 };
  52 static const WCHAR FROM_W[] = { 'F','R','O','M',0 };
  53 static const WCHAR HOLD_W[] = { 'H','O','L','D',0 };
  54 static const WCHAR INSERT_W[] = { 'I','N','S','E','R','T',0 };
  55 static const WCHAR INT_W[] = { 'I','N','T',0 };
  56 static const WCHAR INTEGER_W[] = { 'I','N','T','E','G','E','R',0 };
  57 static const WCHAR INTO_W[] = { 'I','N','T','O',0 };
  58 static const WCHAR IS_W[] = { 'I','S',0 };
  59 static const WCHAR KEY_W[] = { 'K','E','Y',0 };
  60 static const WCHAR LIKE_W[] = { 'L','I','K','E',0 };
  61 static const WCHAR LOCALIZABLE_W[] = { 'L','O','C','A','L','I','Z','A','B','L','E',0 };
  62 static const WCHAR LONG_W[] = { 'L','O','N','G',0 };
  63 static const WCHAR LONGCHAR_W[] = { 'L','O','N','G','C','H','A','R',0 };
  64 static const WCHAR NOT_W[] = { 'N','O','T',0 };
  65 static const WCHAR NULL_W[] = { 'N','U','L','L',0 };
  66 static const WCHAR OBJECT_W[] = { 'O','B','J','E','C','T',0 };
  67 static const WCHAR OR_W[] = { 'O','R',0 };
  68 static const WCHAR ORDER_W[] = { 'O','R','D','E','R',0 };
  69 static const WCHAR PRIMARY_W[] = { 'P','R','I','M','A','R','Y',0 };
  70 static const WCHAR SELECT_W[] = { 'S','E','L','E','C','T',0 };
  71 static const WCHAR SET_W[] = { 'S','E','T',0 };
  72 static const WCHAR SHORT_W[] = { 'S','H','O','R','T',0 };
  73 static const WCHAR TABLE_W[] = { 'T','A','B','L','E',0 };
  74 static const WCHAR TEMPORARY_W[] = { 'T','E','M','P','O','R','A','R','Y',0 };
  75 static const WCHAR UPDATE_W[] = { 'U','P','D','A','T','E',0 };
  76 static const WCHAR VALUES_W[] = { 'V','A','L','U','E','S',0 };
  77 static const WCHAR WHERE_W[] = { 'W','H','E','R','E',0 };
  78
  79 /*
  80 ** These are the keywords
  81 ** They MUST be in alphabetical order
  82 */
  83 static const Keyword aKeywordTable[] = {
  84   { ADD_W, TK_ADD },
  85   { ALTER_W, TK_ALTER },
  86   { AND_W, TK_AND },
  87   { BY_W, TK_BY },
  88   { CHAR_W, TK_CHAR },
  89   { CHARACTER_W, TK_CHAR },
  90   { CREATE_W, TK_CREATE },
  91   { DELETE_W, TK_DELETE },
  92   { DISTINCT_W, TK_DISTINCT },
  93   { DROP_W, TK_DROP },
  94   { FREE_W, TK_FREE },
  95   { FROM_W, TK_FROM },
  96   { HOLD_W, TK_HOLD },
  97   { INSERT_W, TK_INSERT },
  98   { INT_W, TK_INT },
  99   { INTEGER_W, TK_INT },
 100   { INTO_W, TK_INTO },
 101   { IS_W, TK_IS },
 102   { KEY_W, TK_KEY },
 103   { LIKE_W, TK_LIKE },
 104   { LOCALIZABLE_W, TK_LOCALIZABLE },
 105   { LONG_W, TK_LONG },
 106   { LONGCHAR_W, TK_LONGCHAR },
 107   { NOT_W, TK_NOT },
 108   { NULL_W, TK_NULL },
 109   { OBJECT_W, TK_OBJECT },
 110   { OR_W, TK_OR },
 111   { ORDER_W, TK_ORDER },
 112   { PRIMARY_W, TK_PRIMARY },
 113   { SELECT_W, TK_SELECT },
 114   { SET_W, TK_SET },
 115   { SHORT_W, TK_SHORT },
 116   { TABLE_W, TK_TABLE },
 117   { TEMPORARY_W, TK_TEMPORARY },
 118   { UPDATE_W, TK_UPDATE },
 119   { VALUES_W, TK_VALUES },
 120   { WHERE_W, TK_WHERE },
 121 };
 122
 123 #define KEYWORD_COUNT ( sizeof aKeywordTable/sizeof (Keyword) )
 124
 125 /*
 126 ** Comparison function for binary search.
 127 */
 128 static int compKeyword(const void *m1, const void *m2){
 129   const Keyword *k1 = m1, *k2 = m2;
 130
 131   return strcmpiW( k1->zName, k2->zName );
 132 }
 133
 134 /*
 135 ** This function looks up an identifier to determine if it is a
 136 ** keyword.  If it is a keyword, the token code of that keyword is
 137 ** returned.  If the input is not a keyword, TK_ID is returned.
 138 */
 139 static int sqliteKeywordCode(const WCHAR *z, int n){
 140   WCHAR str[MAX_TOKEN_LEN+1];
 141   Keyword key, *r;
 142
 143   if( n>MAX_TOKEN_LEN )
 144     return TK_ID;
 145
 146   memcpy( str, z, n*sizeof (WCHAR) );
 147   str[n] = 0;
 148   key.tokenType = 0;
 149   key.zName = str;
 150   r = bsearch( &key, aKeywordTable, KEYWORD_COUNT, sizeof (Keyword), compKeyword );
 151   if( r )
 152     return r->tokenType;
 153   return TK_ID;
 154 }
 155
 156
 157 /*
 158 ** If X is a character that can be used in an identifier then
 159 ** isIdChar[X] will be 1.  Otherwise isIdChar[X] will be 0.
 160 **
 161 ** In this implementation, an identifier can be a string of
 162 ** alphabetic characters, digits, and "_" plus any character
 163 ** with the high-order bit set.  The latter rule means that
 164 ** any sequence of UTF-8 characters or characters taken from
 165 ** an extended ISO8859 character set can form an identifier.
 166 */
 167 static const char isIdChar[] = {
 168 /* x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 xA xB xC xD xE xF */
 169     0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x */
 170     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 1x */
 171     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,  /* 2x */
 172     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0,  /* 3x */
 173     0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 4x */
 174     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1,  /* 5x */
 175     0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 6x */
 176     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,  /* 7x */
 177     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 8x */
 178     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 9x */
 179     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Ax */
 180     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Bx */
 181     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Cx */
 182     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Dx */
 183     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Ex */
 184     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Fx */
 185 };
 186
 187
 188 /*
 189 ** Return the length of the token that begins at z[0].  Return
 190 ** -1 if the token is (or might be) incomplete.  Store the token
 191 ** type in *tokenType before returning.
 192 */
 193 int sqliteGetToken(const WCHAR *z, int *tokenType, int *skip){
 194   int i;
 195
 196   *skip = 0;
 197   switch( *z ){
 198     case ' ': case '\t': case '\n': case '\f':
 199       for(i=1; isspace(z[i]) && z[i] != '\r'; i++){}
 200       *tokenType = TK_SPACE;
 201       return i;
 202     case '-':
 203       if( z[1]==0 ) return -1;
 204       *tokenType = TK_MINUS;
 205       return 1;
 206     case '(':
 207       *tokenType = TK_LP;
 208       return 1;
 209     case ')':
 210       *tokenType = TK_RP;
 211       return 1;
 212     case '*':
 213       *tokenType = TK_STAR;
 214       return 1;
 215     case '=':
 216       *tokenType = TK_EQ;
 217       return 1;
 218     case '<':
 219       if( z[1]=='=' ){
 220         *tokenType = TK_LE;
 221         return 2;
 222       }else if( z[1]=='>' ){
 223         *tokenType = TK_NE;
 224         return 2;
 225       }else{
 226         *tokenType = TK_LT;
 227         return 1;
 228       }
 229     case '>':
 230       if( z[1]=='=' ){
 231         *tokenType = TK_GE;
 232         return 2;
 233       }else{
 234         *tokenType = TK_GT;
 235         return 1;
 236       }
 237     case '!':
 238       if( z[1]!='=' ){
 239         *tokenType = TK_ILLEGAL;
 240         return 2;
 241       }else{
 242         *tokenType = TK_NE;
 243         return 2;
 244       }
 245     case '?':
 246       *tokenType = TK_WILDCARD;
 247       return 1;
 248     case ',':
 249       *tokenType = TK_COMMA;
 250       return 1;
 251     case '`': case '\'': {
 252       int delim = z[0];
 253       for(i=1; z[i]; i++){
 254         if( z[i]==delim )
 255           break;
 256       }
 257       if( z[i] ) i++;
 258       if( delim == '`' )
 259         *tokenType = TK_ID;
 260       else
 261         *tokenType = TK_STRING;
 262       return i;
 263     }
 264     case '.':
 265       if( !isdigit(z[1]) ){
 266         *tokenType = TK_DOT;
 267         return 1;
 268       }
 269       /* Fall through */
 270     case '0': case '1': case '2': case '3': case '4':
 271     case '5': case '6': case '7': case '8': case '9':
 272       *tokenType = TK_INTEGER;
 273       for(i=1; isdigit(z[i]); i++){}
 274       return i;
 275     case '[':
 276       for(i=1; z[i] && z[i-1]!=']'; i++){}
 277       *tokenType = TK_ID;
 278       return i;
 279     default:
 280       if( !isIdChar[*z] ){
 281         break;
 282       }
 283       for(i=1; isIdChar[z[i]]; i++){}
 284       *tokenType = sqliteKeywordCode(z, i);
 285       if( *tokenType == TK_ID && z[i] == '`' ) *skip = 1;
 286       return i;
 287   }
 288   *tokenType = TK_ILLEGAL;
 289   return 1;
 290 }