xref: /reactos/dll/win32/msi/tokenize.c (revision c7bba39a)
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 *name;             /* The keyword name */
36   unsigned int len;
37   int tokenType;           /* The token value for this keyword */
38 };
39 
40 #define MAX_TOKEN_LEN 11
41 
42 static const WCHAR addW[] = {'A','D','D'};
43 static const WCHAR alterW[] = {'A','L','T','E','R'};
44 static const WCHAR andW[] = {'A','N','D'};
45 static const WCHAR byW[] = {'B','Y'};
46 static const WCHAR charW[] = {'C','H','A','R'};
47 static const WCHAR characterW[] = {'C','H','A','R','A','C','T','E','R'};
48 static const WCHAR createW[] = {'C','R','E','A','T','E'};
49 static const WCHAR deleteW[] = {'D','E','L','E','T','E'};
50 static const WCHAR distinctW[] = {'D','I','S','T','I','N','C','T'};
51 static const WCHAR dropW[] = {'D','R','O','P'};
52 static const WCHAR freeW[] = {'F','R','E','E'};
53 static const WCHAR fromW[] = {'F','R','O','M'};
54 static const WCHAR holdW[] = {'H','O','L','D'};
55 static const WCHAR insertW[] = {'I','N','S','E','R','T'};
56 static const WCHAR intW[] = {'I','N','T'};
57 static const WCHAR integerW[] = {'I','N','T','E','G','E','R'};
58 static const WCHAR intoW[] = {'I','N','T','O'};
59 static const WCHAR isW[] = {'I','S'};
60 static const WCHAR keyW[] = {'K','E','Y'};
61 static const WCHAR likeW[] = {'L','I','K','E'};
62 static const WCHAR localizableW[] = {'L','O','C','A','L','I','Z','A','B','L','E'};
63 static const WCHAR longW[] = {'L','O','N','G'};
64 static const WCHAR longcharW[] = {'L','O','N','G','C','H','A','R'};
65 static const WCHAR notW[] = {'N','O','T'};
66 static const WCHAR nullW[] = {'N','U','L','L'};
67 static const WCHAR objectW[] = {'O','B','J','E','C','T'};
68 static const WCHAR orW[] = {'O','R'};
69 static const WCHAR orderW[] = {'O','R','D','E','R'};
70 static const WCHAR primaryW[] = {'P','R','I','M','A','R','Y'};
71 static const WCHAR selectW[] = {'S','E','L','E','C','T'};
72 static const WCHAR setW[] = {'S','E','T'};
73 static const WCHAR shortW[] = {'S','H','O','R','T'};
74 static const WCHAR tableW[] = {'T','A','B','L','E'};
75 static const WCHAR temporaryW[] = {'T','E','M','P','O','R','A','R','Y'};
76 static const WCHAR updateW[] = {'U','P','D','A','T','E'};
77 static const WCHAR valuesW[] = {'V','A','L','U','E','S'};
78 static const WCHAR whereW[] = {'W','H','E','R','E'};
79 
80 #define ARRAY_SIZE(array) (sizeof(array)/sizeof((array)[0]))
81 
82 /*
83 ** These are the keywords
84 ** They MUST be in alphabetical order
85 */
86 static const Keyword aKeywordTable[] = {
87   { addW,         ARRAY_SIZE(addW),         TK_ADD },
88   { alterW,       ARRAY_SIZE(alterW),       TK_ALTER },
89   { andW,         ARRAY_SIZE(andW),         TK_AND },
90   { byW,          ARRAY_SIZE(byW),          TK_BY },
91   { charW,        ARRAY_SIZE(charW),        TK_CHAR },
92   { characterW,   ARRAY_SIZE(characterW),   TK_CHAR },
93   { createW,      ARRAY_SIZE(createW),      TK_CREATE },
94   { deleteW,      ARRAY_SIZE(deleteW),      TK_DELETE },
95   { distinctW,    ARRAY_SIZE(distinctW),    TK_DISTINCT },
96   { dropW,        ARRAY_SIZE(dropW),        TK_DROP },
97   { freeW,        ARRAY_SIZE(freeW),        TK_FREE },
98   { fromW,        ARRAY_SIZE(fromW),        TK_FROM },
99   { holdW,        ARRAY_SIZE(holdW),        TK_HOLD },
100   { insertW,      ARRAY_SIZE(insertW),      TK_INSERT },
101   { intW,         ARRAY_SIZE(intW),         TK_INT },
102   { integerW,     ARRAY_SIZE(integerW),     TK_INT },
103   { intoW,        ARRAY_SIZE(intoW),        TK_INTO },
104   { isW,          ARRAY_SIZE(isW),          TK_IS },
105   { keyW,         ARRAY_SIZE(keyW),         TK_KEY },
106   { likeW,        ARRAY_SIZE(likeW),        TK_LIKE },
107   { localizableW, ARRAY_SIZE(localizableW), TK_LOCALIZABLE },
108   { longW,        ARRAY_SIZE(longW),        TK_LONG },
109   { longcharW,    ARRAY_SIZE(longcharW),    TK_LONGCHAR },
110   { notW,         ARRAY_SIZE(notW),         TK_NOT },
111   { nullW,        ARRAY_SIZE(nullW),        TK_NULL },
112   { objectW,      ARRAY_SIZE(objectW),      TK_OBJECT },
113   { orW,          ARRAY_SIZE(orW),          TK_OR },
114   { orderW,       ARRAY_SIZE(orderW),       TK_ORDER },
115   { primaryW,     ARRAY_SIZE(primaryW),     TK_PRIMARY },
116   { selectW,      ARRAY_SIZE(selectW),      TK_SELECT },
117   { setW,         ARRAY_SIZE(setW),         TK_SET },
118   { shortW,       ARRAY_SIZE(shortW),       TK_SHORT },
119   { tableW,       ARRAY_SIZE(tableW),       TK_TABLE },
120   { temporaryW,   ARRAY_SIZE(temporaryW),   TK_TEMPORARY },
121   { updateW,      ARRAY_SIZE(updateW),      TK_UPDATE },
122   { valuesW,      ARRAY_SIZE(valuesW),      TK_VALUES },
123   { whereW,       ARRAY_SIZE(whereW),       TK_WHERE },
124 };
125 
126 /*
127 ** Comparison function for binary search.
128 */
129 static int compKeyword(const void *m1, const void *m2){
130   const Keyword *k1 = m1, *k2 = m2;
131   int ret, len = min( k1->len, k2->len );
132 
133   if ((ret = memicmpW( k1->name, k2->name, len ))) return ret;
134   if (k1->len < k2->len) return -1;
135   else if (k1->len > k2->len) return 1;
136   return 0;
137 }
138 
139 /*
140 ** This function looks up an identifier to determine if it is a
141 ** keyword.  If it is a keyword, the token code of that keyword is
142 ** returned.  If the input is not a keyword, TK_ID is returned.
143 */
144 static int sqliteKeywordCode(const WCHAR *z, int n){
145   Keyword key, *r;
146 
147   if( n>MAX_TOKEN_LEN )
148     return TK_ID;
149 
150   key.tokenType = 0;
151   key.name = z;
152   key.len = n;
153   r = bsearch( &key, aKeywordTable, ARRAY_SIZE(aKeywordTable), sizeof(Keyword), compKeyword );
154   if( r )
155     return r->tokenType;
156   return TK_ID;
157 }
158 
159 
160 /*
161 ** If X is a character that can be used in an identifier then
162 ** isIdChar[X] will be 1.  Otherwise isIdChar[X] will be 0.
163 **
164 ** In this implementation, an identifier can be a string of
165 ** alphabetic characters, digits, and "_" plus any character
166 ** with the high-order bit set.  The latter rule means that
167 ** any sequence of UTF-8 characters or characters taken from
168 ** an extended ISO8859 character set can form an identifier.
169 */
170 static const char isIdChar[] = {
171 /* x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 xA xB xC xD xE xF */
172     0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x */
173     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 1x */
174     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,  /* 2x */
175     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0,  /* 3x */
176     0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 4x */
177     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1,  /* 5x */
178     0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 6x */
179     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,  /* 7x */
180     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 8x */
181     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 9x */
182     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Ax */
183     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Bx */
184     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Cx */
185     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Dx */
186     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Ex */
187     1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* Fx */
188 };
189 
190 /*
191 ** WCHAR safe version of isdigit()
192 */
193 static inline int isDigit(WCHAR c)
194 {
195     return c >= '0' && c <= '9';
196 }
197 
198 /*
199 ** WCHAR safe version of isspace(), except '\r'
200 */
201 static inline int isSpace(WCHAR c)
202 {
203     return c == ' ' || c == '\t' || c == '\n' || c == '\f';
204 }
205 
206 /*
207 ** Return the length of the token that begins at z[0].  Return
208 ** -1 if the token is (or might be) incomplete.  Store the token
209 ** type in *tokenType before returning.
210 */
211 int sqliteGetToken(const WCHAR *z, int *tokenType, int *skip){
212   int i;
213 
214   *skip = 0;
215   switch( *z ){
216     case ' ': case '\t': case '\n': case '\f':
217       for(i=1; isSpace(z[i]); i++){}
218       *tokenType = TK_SPACE;
219       return i;
220     case '-':
221       if( z[1]==0 ) return -1;
222       *tokenType = TK_MINUS;
223       return 1;
224     case '(':
225       *tokenType = TK_LP;
226       return 1;
227     case ')':
228       *tokenType = TK_RP;
229       return 1;
230     case '*':
231       *tokenType = TK_STAR;
232       return 1;
233     case '=':
234       *tokenType = TK_EQ;
235       return 1;
236     case '<':
237       if( z[1]=='=' ){
238         *tokenType = TK_LE;
239         return 2;
240       }else if( z[1]=='>' ){
241         *tokenType = TK_NE;
242         return 2;
243       }else{
244         *tokenType = TK_LT;
245         return 1;
246       }
247     case '>':
248       if( z[1]=='=' ){
249         *tokenType = TK_GE;
250         return 2;
251       }else{
252         *tokenType = TK_GT;
253         return 1;
254       }
255     case '!':
256       if( z[1]!='=' ){
257         *tokenType = TK_ILLEGAL;
258         return 2;
259       }else{
260         *tokenType = TK_NE;
261         return 2;
262       }
263     case '?':
264       *tokenType = TK_WILDCARD;
265       return 1;
266     case ',':
267       *tokenType = TK_COMMA;
268       return 1;
269     case '`': case '\'': {
270       int delim = z[0];
271       for(i=1; z[i]; i++){
272         if( z[i]==delim )
273           break;
274       }
275       if( z[i] ) i++;
276       if( delim == '`' )
277         *tokenType = TK_ID;
278       else
279         *tokenType = TK_STRING;
280       return i;
281     }
282     case '.':
283       if( !isDigit(z[1]) ){
284         *tokenType = TK_DOT;
285         return 1;
286       }
287       /* Fall through */
288     case '0': case '1': case '2': case '3': case '4':
289     case '5': case '6': case '7': case '8': case '9':
290       *tokenType = TK_INTEGER;
291       for(i=1; isDigit(z[i]); i++){}
292       return i;
293     case '[':
294       for(i=1; z[i] && z[i-1]!=']'; i++){}
295       *tokenType = TK_ID;
296       return i;
297     default:
298       if( !isIdChar[*z] ){
299         break;
300       }
301       for(i=1; isIdChar[z[i]]; i++){}
302       *tokenType = sqliteKeywordCode(z, i);
303       if( *tokenType == TK_ID && z[i] == '`' ) *skip = 1;
304       return i;
305   }
306   *tokenType = TK_ILLEGAL;
307   return 1;
308 }
309