xref: /reactos/dll/win32/vbscript/lex.c (revision 7eead935)
1 /*
2  * Copyright 2011 Jacek Caban for CodeWeavers
3  *
4  * This library is free software; you can redistribute it and/or
5  * modify it under the terms of the GNU Lesser General Public
6  * License as published by the Free Software Foundation; either
7  * version 2.1 of the License, or (at your option) any later version.
8  *
9  * This library 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  * Lesser General Public License for more details.
13  *
14  * You should have received a copy of the GNU Lesser General Public
15  * License along with this library; if not, write to the Free Software
16  * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
17  */
18 
19 #include "config.h"
20 #include "wine/port.h"
21 
22 #include <assert.h>
23 #include <limits.h>
24 
25 #include "vbscript.h"
26 #include "parse.h"
27 #include "parser.tab.h"
28 
29 #include "wine/debug.h"
30 
31 WINE_DEFAULT_DEBUG_CHANNEL(vbscript);
32 
33 static const WCHAR andW[] = {'a','n','d',0};
34 static const WCHAR byrefW[] = {'b','y','r','e','f',0};
35 static const WCHAR byvalW[] = {'b','y','v','a','l',0};
36 static const WCHAR callW[] = {'c','a','l','l',0};
37 static const WCHAR caseW[] = {'c','a','s','e',0};
38 static const WCHAR classW[] = {'c','l','a','s','s',0};
39 static const WCHAR constW[] = {'c','o','n','s','t',0};
40 static const WCHAR defaultW[] = {'d','e','f','a','u','l','t',0};
41 static const WCHAR dimW[] = {'d','i','m',0};
42 static const WCHAR doW[] = {'d','o',0};
43 static const WCHAR eachW[] = {'e','a','c','h',0};
44 static const WCHAR elseW[] = {'e','l','s','e',0};
45 static const WCHAR elseifW[] = {'e','l','s','e','i','f',0};
46 static const WCHAR emptyW[] = {'e','m','p','t','y',0};
47 static const WCHAR endW[] = {'e','n','d',0};
48 static const WCHAR eqvW[] = {'e','q','v',0};
49 static const WCHAR errorW[] = {'e','r','r','o','r',0};
50 static const WCHAR exitW[] = {'e','x','i','t',0};
51 static const WCHAR explicitW[] = {'e','x','p','l','i','c','i','t',0};
52 static const WCHAR falseW[] = {'f','a','l','s','e',0};
53 static const WCHAR forW[] = {'f','o','r',0};
54 static const WCHAR functionW[] = {'f','u','n','c','t','i','o','n',0};
55 static const WCHAR getW[] = {'g','e','t',0};
56 static const WCHAR gotoW[] = {'g','o','t','o',0};
57 static const WCHAR ifW[] = {'i','f',0};
58 static const WCHAR impW[] = {'i','m','p',0};
59 static const WCHAR inW[] = {'i','n',0};
60 static const WCHAR isW[] = {'i','s',0};
61 static const WCHAR letW[] = {'l','e','t',0};
62 static const WCHAR loopW[] = {'l','o','o','p',0};
63 static const WCHAR meW[] = {'m','e',0};
64 static const WCHAR modW[] = {'m','o','d',0};
65 static const WCHAR newW[] = {'n','e','w',0};
66 static const WCHAR nextW[] = {'n','e','x','t',0};
67 static const WCHAR notW[] = {'n','o','t',0};
68 static const WCHAR nothingW[] = {'n','o','t','h','i','n','g',0};
69 static const WCHAR nullW[] = {'n','u','l','l',0};
70 static const WCHAR onW[] = {'o','n',0};
71 static const WCHAR optionW[] = {'o','p','t','i','o','n',0};
72 static const WCHAR orW[] = {'o','r',0};
73 static const WCHAR privateW[] = {'p','r','i','v','a','t','e',0};
74 static const WCHAR propertyW[] = {'p','r','o','p','e','r','t','y',0};
75 static const WCHAR publicW[] = {'p','u','b','l','i','c',0};
76 static const WCHAR remW[] = {'r','e','m',0};
77 static const WCHAR resumeW[] = {'r','e','s','u','m','e',0};
78 static const WCHAR selectW[] = {'s','e','l','e','c','t',0};
79 static const WCHAR setW[] = {'s','e','t',0};
80 static const WCHAR stepW[] = {'s','t','e','p',0};
81 static const WCHAR stopW[] = {'s','t','o','p',0};
82 static const WCHAR subW[] = {'s','u','b',0};
83 static const WCHAR thenW[] = {'t','h','e','n',0};
84 static const WCHAR toW[] = {'t','o',0};
85 static const WCHAR trueW[] = {'t','r','u','e',0};
86 static const WCHAR untilW[] = {'u','n','t','i','l',0};
87 static const WCHAR wendW[] = {'w','e','n','d',0};
88 static const WCHAR whileW[] = {'w','h','i','l','e',0};
89 static const WCHAR xorW[] = {'x','o','r',0};
90 
91 static const struct {
92     const WCHAR *word;
93     int token;
94 } keywords[] = {
95     {andW,       tAND},
96     {byrefW,     tBYREF},
97     {byvalW,     tBYVAL},
98     {callW,      tCALL},
99     {caseW,      tCASE},
100     {classW,     tCLASS},
101     {constW,     tCONST},
102     {defaultW,   tDEFAULT},
103     {dimW,       tDIM},
104     {doW,        tDO},
105     {eachW,      tEACH},
106     {elseW,      tELSE},
107     {elseifW,    tELSEIF},
108     {emptyW,     tEMPTY},
109     {endW,       tEND},
110     {eqvW,       tEQV},
111     {errorW,     tERROR},
112     {exitW,      tEXIT},
113     {explicitW,  tEXPLICIT},
114     {falseW,     tFALSE},
115     {forW,       tFOR},
116     {functionW,  tFUNCTION},
117     {getW,       tGET},
118     {gotoW,      tGOTO},
119     {ifW,        tIF},
120     {impW,       tIMP},
121     {inW,        tIN},
122     {isW,        tIS},
123     {letW,       tLET},
124     {loopW,      tLOOP},
125     {meW,        tME},
126     {modW,       tMOD},
127     {newW,       tNEW},
128     {nextW,      tNEXT},
129     {notW,       tNOT},
130     {nothingW,   tNOTHING},
131     {nullW,      tNULL},
132     {onW,        tON},
133     {optionW,    tOPTION},
134     {orW,        tOR},
135     {privateW,   tPRIVATE},
136     {propertyW,  tPROPERTY},
137     {publicW,    tPUBLIC},
138     {remW,       tREM},
139     {resumeW,    tRESUME},
140     {selectW,    tSELECT},
141     {setW,       tSET},
142     {stepW,      tSTEP},
143     {stopW,      tSTOP},
144     {subW,       tSUB},
145     {thenW,      tTHEN},
146     {toW,        tTO},
147     {trueW,      tTRUE},
148     {untilW,     tUNTIL},
149     {wendW,      tWEND},
150     {whileW,     tWHILE},
151     {xorW,       tXOR}
152 };
153 
154 static inline BOOL is_identifier_char(WCHAR c)
155 {
156     return isalnumW(c) || c == '_';
157 }
158 
159 static int check_keyword(parser_ctx_t *ctx, const WCHAR *word)
160 {
161     const WCHAR *p1 = ctx->ptr;
162     const WCHAR *p2 = word;
163     WCHAR c;
164 
165     while(p1 < ctx->end && *p2) {
166         c = tolowerW(*p1);
167         if(c != *p2)
168             return c - *p2;
169         p1++;
170         p2++;
171     }
172 
173     if(*p2 || (p1 < ctx->end && is_identifier_char(*p1)))
174         return 1;
175 
176     ctx->ptr = p1;
177     return 0;
178 }
179 
180 static int check_keywords(parser_ctx_t *ctx)
181 {
182     int min = 0, max = ARRAY_SIZE(keywords)-1, r, i;
183 
184     while(min <= max) {
185         i = (min+max)/2;
186 
187         r = check_keyword(ctx, keywords[i].word);
188         if(!r)
189             return keywords[i].token;
190 
191         if(r > 0)
192             min = i+1;
193         else
194             max = i-1;
195     }
196 
197     return 0;
198 }
199 
200 static int parse_identifier(parser_ctx_t *ctx, const WCHAR **ret)
201 {
202     const WCHAR *ptr = ctx->ptr++;
203     WCHAR *str;
204     int len;
205 
206     while(ctx->ptr < ctx->end && is_identifier_char(*ctx->ptr))
207         ctx->ptr++;
208     len = ctx->ptr-ptr;
209 
210     str = parser_alloc(ctx, (len+1)*sizeof(WCHAR));
211     if(!str)
212         return 0;
213 
214     memcpy(str, ptr, (len+1)*sizeof(WCHAR));
215     str[len] = 0;
216     *ret = str;
217     return tIdentifier;
218 }
219 
220 static int parse_string_literal(parser_ctx_t *ctx, const WCHAR **ret)
221 {
222     const WCHAR *ptr = ++ctx->ptr;
223     WCHAR *rptr;
224     int len = 0;
225 
226     while(ctx->ptr < ctx->end) {
227         if(*ctx->ptr == '\n') {
228             FIXME("newline inside string literal\n");
229             return 0;
230         }
231 
232        if(*ctx->ptr == '"') {
233             if(ctx->ptr[1] != '"')
234                 break;
235             len--;
236             ctx->ptr++;
237         }
238         ctx->ptr++;
239     }
240 
241     if(ctx->ptr == ctx->end) {
242         FIXME("unterminated string literal\n");
243         return 0;
244     }
245 
246     len += ctx->ptr-ptr;
247 
248     *ret = rptr = parser_alloc(ctx, (len+1)*sizeof(WCHAR));
249     if(!rptr)
250         return 0;
251 
252     while(ptr < ctx->ptr) {
253         if(*ptr == '"')
254             ptr++;
255         *rptr++ = *ptr++;
256     }
257 
258     *rptr = 0;
259     ctx->ptr++;
260     return tString;
261 }
262 
263 static int parse_numeric_literal(parser_ctx_t *ctx, void **ret)
264 {
265     BOOL use_int = TRUE;
266     LONGLONG d = 0, hlp;
267     int exp = 0;
268     double r;
269 
270     if(*ctx->ptr == '0' && !('0' <= ctx->ptr[1] && ctx->ptr[1] <= '9') && ctx->ptr[1] != '.')
271         return *ctx->ptr++;
272 
273     while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr)) {
274         hlp = d*10 + *(ctx->ptr++) - '0';
275         if(d>MAXLONGLONG/10 || hlp<0) {
276             exp++;
277             break;
278         }
279         else
280             d = hlp;
281     }
282     while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr)) {
283         exp++;
284         ctx->ptr++;
285     }
286 
287     if(*ctx->ptr == '.') {
288         use_int = FALSE;
289         ctx->ptr++;
290 
291         while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr)) {
292             hlp = d*10 + *(ctx->ptr++) - '0';
293             if(d>MAXLONGLONG/10 || hlp<0)
294                 break;
295 
296             d = hlp;
297             exp--;
298         }
299         while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr))
300             ctx->ptr++;
301     }
302 
303     if(*ctx->ptr == 'e' || *ctx->ptr == 'E') {
304         int e = 0, sign = 1;
305 
306         if(*++ctx->ptr == '-') {
307             ctx->ptr++;
308             sign = -1;
309         }
310 
311         if(!isdigitW(*ctx->ptr)) {
312             FIXME("Invalid numeric literal\n");
313             return 0;
314         }
315 
316         use_int = FALSE;
317 
318         do {
319             e = e*10 + *(ctx->ptr++) - '0';
320             if(sign == -1 && -e+exp < -(INT_MAX/100)) {
321                 /* The literal will be rounded to 0 anyway. */
322                 while(isdigitW(*ctx->ptr))
323                     ctx->ptr++;
324                 *(double*)ret = 0;
325                 return tDouble;
326             }
327 
328             if(sign*e + exp > INT_MAX/100) {
329                 FIXME("Invalid numeric literal\n");
330                 return 0;
331             }
332         } while(isdigitW(*ctx->ptr));
333 
334         exp += sign*e;
335     }
336 
337     if(use_int && (LONG)d == d) {
338         LONG l = d;
339         *(LONG*)ret = l;
340         return (short)l == l ? tShort : tLong;
341     }
342 
343     r = exp>=0 ? d*pow(10, exp) : d/pow(10, -exp);
344     if(isinf(r)) {
345         FIXME("Invalid numeric literal\n");
346         return 0;
347     }
348 
349     *(double*)ret = r;
350     return tDouble;
351 }
352 
353 static int hex_to_int(WCHAR c)
354 {
355     if('0' <= c && c <= '9')
356         return c-'0';
357     if('a' <= c && c <= 'f')
358         return c+10-'a';
359     if('A' <= c && c <= 'F')
360         return c+10-'A';
361     return -1;
362 }
363 
364 static int parse_hex_literal(parser_ctx_t *ctx, LONG *ret)
365 {
366     const WCHAR *begin = ctx->ptr;
367     LONG l = 0, d;
368 
369     while((d = hex_to_int(*++ctx->ptr)) != -1)
370         l = l*16 + d;
371 
372     if(begin + 9 /* max digits+1 */ < ctx->ptr || (*ctx->ptr != '&' && is_identifier_char(*ctx->ptr))) {
373         FIXME("invalid literal\n");
374         return 0;
375     }
376 
377     if(*ctx->ptr == '&')
378         ctx->ptr++;
379 
380     *ret = l;
381     return (short)l == l ? tShort : tLong;
382 }
383 
384 static void skip_spaces(parser_ctx_t *ctx)
385 {
386     while(*ctx->ptr == ' ' || *ctx->ptr == '\t')
387         ctx->ptr++;
388 }
389 
390 static int comment_line(parser_ctx_t *ctx)
391 {
392     static const WCHAR newlineW[] = {'\n','\r',0};
393     ctx->ptr = strpbrkW(ctx->ptr, newlineW);
394     if(ctx->ptr)
395         ctx->ptr++;
396     else
397         ctx->ptr = ctx->end;
398     return tNL;
399 }
400 
401 static int parse_next_token(void *lval, parser_ctx_t *ctx)
402 {
403     WCHAR c;
404 
405     skip_spaces(ctx);
406     if(ctx->ptr == ctx->end)
407         return ctx->last_token == tNL ? tEOF : tNL;
408 
409     c = *ctx->ptr;
410 
411     if('0' <= c && c <= '9')
412         return parse_numeric_literal(ctx, lval);
413 
414     if(isalphaW(c)) {
415         int ret = check_keywords(ctx);
416         if(!ret)
417             return parse_identifier(ctx, lval);
418         if(ret != tREM)
419             return ret;
420         c = '\'';
421     }
422 
423     switch(c) {
424     case '\n':
425     case '\r':
426         ctx->ptr++;
427         return tNL;
428     case '\'':
429         return comment_line(ctx);
430     case ':':
431     case ')':
432     case ',':
433     case '=':
434     case '+':
435     case '*':
436     case '/':
437     case '^':
438     case '\\':
439     case '.':
440     case '_':
441         return *ctx->ptr++;
442     case '-':
443         if(ctx->is_html && ctx->ptr[1] == '-' && ctx->ptr[2] == '>')
444             return comment_line(ctx);
445         ctx->ptr++;
446         return '-';
447     case '(':
448         /* NOTE:
449          * We resolve empty brackets in lexer instead of parser to avoid complex conflicts
450          * in call statement special case |f()| without 'call' keyword
451          */
452         ctx->ptr++;
453         skip_spaces(ctx);
454         if(*ctx->ptr == ')') {
455             ctx->ptr++;
456             return tEMPTYBRACKETS;
457         }
458         return '(';
459     case '"':
460         return parse_string_literal(ctx, lval);
461     case '&':
462         if(*++ctx->ptr == 'h' || *ctx->ptr == 'H')
463             return parse_hex_literal(ctx, lval);
464         return '&';
465     case '<':
466         switch(*++ctx->ptr) {
467         case '>':
468             ctx->ptr++;
469             return tNEQ;
470         case '=':
471             ctx->ptr++;
472             return tLTEQ;
473         case '!':
474             if(ctx->is_html && ctx->ptr[1] == '-' && ctx->ptr[2] == '-')
475                 return comment_line(ctx);
476         }
477         return '<';
478     case '>':
479         if(*++ctx->ptr == '=') {
480             ctx->ptr++;
481             return tGTEQ;
482         }
483         return '>';
484     default:
485         FIXME("Unhandled char %c in %s\n", *ctx->ptr, debugstr_w(ctx->ptr));
486     }
487 
488     return 0;
489 }
490 
491 int parser_lex(void *lval, parser_ctx_t *ctx)
492 {
493     int ret;
494 
495     while(1) {
496         ret = parse_next_token(lval, ctx);
497         if(ret == '_') {
498             skip_spaces(ctx);
499             if(*ctx->ptr != '\n') {
500                 FIXME("'_' not followed by newline\n");
501                 return 0;
502             }
503             ctx->ptr++;
504             continue;
505         }
506         if(ret != tNL || ctx->last_token != tNL)
507             break;
508 
509         ctx->last_nl = ctx->ptr-ctx->code;
510     }
511 
512     return (ctx->last_token = ret);
513 }
514