xref: /reactos/dll/win32/vbscript/lex.c (revision 53221834)
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 #ifdef __REACTOS__
20 #include <wine/config.h>
21 #include <wine/port.h>
22 #endif
23 #include <assert.h>
24 #include <limits.h>
25 #include <math.h>
26 
27 #include "vbscript.h"
28 #include "parse.h"
29 #include "parser.tab.h"
30 
31 #include "wine/debug.h"
32 
33 WINE_DEFAULT_DEBUG_CHANNEL(vbscript);
34 
35 static const WCHAR andW[] = {'a','n','d',0};
36 static const WCHAR byrefW[] = {'b','y','r','e','f',0};
37 static const WCHAR byvalW[] = {'b','y','v','a','l',0};
38 static const WCHAR callW[] = {'c','a','l','l',0};
39 static const WCHAR caseW[] = {'c','a','s','e',0};
40 static const WCHAR classW[] = {'c','l','a','s','s',0};
41 static const WCHAR constW[] = {'c','o','n','s','t',0};
42 static const WCHAR defaultW[] = {'d','e','f','a','u','l','t',0};
43 static const WCHAR dimW[] = {'d','i','m',0};
44 static const WCHAR doW[] = {'d','o',0};
45 static const WCHAR eachW[] = {'e','a','c','h',0};
46 static const WCHAR elseW[] = {'e','l','s','e',0};
47 static const WCHAR elseifW[] = {'e','l','s','e','i','f',0};
48 static const WCHAR emptyW[] = {'e','m','p','t','y',0};
49 static const WCHAR endW[] = {'e','n','d',0};
50 static const WCHAR eqvW[] = {'e','q','v',0};
51 static const WCHAR errorW[] = {'e','r','r','o','r',0};
52 static const WCHAR exitW[] = {'e','x','i','t',0};
53 static const WCHAR explicitW[] = {'e','x','p','l','i','c','i','t',0};
54 static const WCHAR falseW[] = {'f','a','l','s','e',0};
55 static const WCHAR forW[] = {'f','o','r',0};
56 static const WCHAR functionW[] = {'f','u','n','c','t','i','o','n',0};
57 static const WCHAR getW[] = {'g','e','t',0};
58 static const WCHAR gotoW[] = {'g','o','t','o',0};
59 static const WCHAR ifW[] = {'i','f',0};
60 static const WCHAR impW[] = {'i','m','p',0};
61 static const WCHAR inW[] = {'i','n',0};
62 static const WCHAR isW[] = {'i','s',0};
63 static const WCHAR letW[] = {'l','e','t',0};
64 static const WCHAR loopW[] = {'l','o','o','p',0};
65 static const WCHAR meW[] = {'m','e',0};
66 static const WCHAR modW[] = {'m','o','d',0};
67 static const WCHAR newW[] = {'n','e','w',0};
68 static const WCHAR nextW[] = {'n','e','x','t',0};
69 static const WCHAR notW[] = {'n','o','t',0};
70 static const WCHAR nothingW[] = {'n','o','t','h','i','n','g',0};
71 static const WCHAR nullW[] = {'n','u','l','l',0};
72 static const WCHAR onW[] = {'o','n',0};
73 static const WCHAR optionW[] = {'o','p','t','i','o','n',0};
74 static const WCHAR orW[] = {'o','r',0};
75 static const WCHAR privateW[] = {'p','r','i','v','a','t','e',0};
76 static const WCHAR propertyW[] = {'p','r','o','p','e','r','t','y',0};
77 static const WCHAR publicW[] = {'p','u','b','l','i','c',0};
78 static const WCHAR remW[] = {'r','e','m',0};
79 static const WCHAR resumeW[] = {'r','e','s','u','m','e',0};
80 static const WCHAR selectW[] = {'s','e','l','e','c','t',0};
81 static const WCHAR setW[] = {'s','e','t',0};
82 static const WCHAR stepW[] = {'s','t','e','p',0};
83 static const WCHAR stopW[] = {'s','t','o','p',0};
84 static const WCHAR subW[] = {'s','u','b',0};
85 static const WCHAR thenW[] = {'t','h','e','n',0};
86 static const WCHAR toW[] = {'t','o',0};
87 static const WCHAR trueW[] = {'t','r','u','e',0};
88 static const WCHAR untilW[] = {'u','n','t','i','l',0};
89 static const WCHAR wendW[] = {'w','e','n','d',0};
90 static const WCHAR whileW[] = {'w','h','i','l','e',0};
91 static const WCHAR xorW[] = {'x','o','r',0};
92 
93 static const struct {
94     const WCHAR *word;
95     int token;
96 } keywords[] = {
97     {andW,       tAND},
98     {byrefW,     tBYREF},
99     {byvalW,     tBYVAL},
100     {callW,      tCALL},
101     {caseW,      tCASE},
102     {classW,     tCLASS},
103     {constW,     tCONST},
104     {defaultW,   tDEFAULT},
105     {dimW,       tDIM},
106     {doW,        tDO},
107     {eachW,      tEACH},
108     {elseW,      tELSE},
109     {elseifW,    tELSEIF},
110     {emptyW,     tEMPTY},
111     {endW,       tEND},
112     {eqvW,       tEQV},
113     {errorW,     tERROR},
114     {exitW,      tEXIT},
115     {explicitW,  tEXPLICIT},
116     {falseW,     tFALSE},
117     {forW,       tFOR},
118     {functionW,  tFUNCTION},
119     {getW,       tGET},
120     {gotoW,      tGOTO},
121     {ifW,        tIF},
122     {impW,       tIMP},
123     {inW,        tIN},
124     {isW,        tIS},
125     {letW,       tLET},
126     {loopW,      tLOOP},
127     {meW,        tME},
128     {modW,       tMOD},
129     {newW,       tNEW},
130     {nextW,      tNEXT},
131     {notW,       tNOT},
132     {nothingW,   tNOTHING},
133     {nullW,      tNULL},
134     {onW,        tON},
135     {optionW,    tOPTION},
136     {orW,        tOR},
137     {privateW,   tPRIVATE},
138     {propertyW,  tPROPERTY},
139     {publicW,    tPUBLIC},
140     {remW,       tREM},
141     {resumeW,    tRESUME},
142     {selectW,    tSELECT},
143     {setW,       tSET},
144     {stepW,      tSTEP},
145     {stopW,      tSTOP},
146     {subW,       tSUB},
147     {thenW,      tTHEN},
148     {toW,        tTO},
149     {trueW,      tTRUE},
150     {untilW,     tUNTIL},
151     {wendW,      tWEND},
152     {whileW,     tWHILE},
153     {xorW,       tXOR}
154 };
155 
156 static inline BOOL is_identifier_char(WCHAR c)
157 {
158     return iswalnum(c) || c == '_';
159 }
160 
161 static int check_keyword(parser_ctx_t *ctx, const WCHAR *word, const WCHAR **lval)
162 {
163     const WCHAR *p1 = ctx->ptr;
164     const WCHAR *p2 = word;
165     WCHAR c;
166 
167     while(p1 < ctx->end && *p2) {
168         c = towlower(*p1);
169         if(c != *p2)
170             return c - *p2;
171         p1++;
172         p2++;
173     }
174 
175     if(*p2 || (p1 < ctx->end && is_identifier_char(*p1)))
176         return 1;
177 
178     ctx->ptr = p1;
179     *lval = word;
180     return 0;
181 }
182 
183 static int check_keywords(parser_ctx_t *ctx, const WCHAR **lval)
184 {
185     int min = 0, max = ARRAY_SIZE(keywords)-1, r, i;
186 
187     while(min <= max) {
188         i = (min+max)/2;
189 
190         r = check_keyword(ctx, keywords[i].word, lval);
191         if(!r)
192             return keywords[i].token;
193 
194         if(r > 0)
195             min = i+1;
196         else
197             max = i-1;
198     }
199 
200     return 0;
201 }
202 
203 static int parse_identifier(parser_ctx_t *ctx, const WCHAR **ret)
204 {
205     const WCHAR *ptr = ctx->ptr++;
206     WCHAR *str;
207     int len;
208 
209     while(ctx->ptr < ctx->end && is_identifier_char(*ctx->ptr))
210         ctx->ptr++;
211     len = ctx->ptr-ptr;
212 
213     str = parser_alloc(ctx, (len+1)*sizeof(WCHAR));
214     if(!str)
215         return 0;
216 
217     memcpy(str, ptr, (len+1)*sizeof(WCHAR));
218     str[len] = 0;
219     *ret = str;
220     return tIdentifier;
221 }
222 
223 static int parse_string_literal(parser_ctx_t *ctx, const WCHAR **ret)
224 {
225     const WCHAR *ptr = ++ctx->ptr;
226     WCHAR *rptr;
227     int len = 0;
228 
229     while(ctx->ptr < ctx->end) {
230         if(*ctx->ptr == '\n' || *ctx->ptr == '\r') {
231             FIXME("newline inside string literal\n");
232             return 0;
233         }
234 
235        if(*ctx->ptr == '"') {
236             if(ctx->ptr[1] != '"')
237                 break;
238             len--;
239             ctx->ptr++;
240         }
241         ctx->ptr++;
242     }
243 
244     if(ctx->ptr == ctx->end) {
245         FIXME("unterminated string literal\n");
246         return 0;
247     }
248 
249     len += ctx->ptr-ptr;
250 
251     *ret = rptr = parser_alloc(ctx, (len+1)*sizeof(WCHAR));
252     if(!rptr)
253         return 0;
254 
255     while(ptr < ctx->ptr) {
256         if(*ptr == '"')
257             ptr++;
258         *rptr++ = *ptr++;
259     }
260 
261     *rptr = 0;
262     ctx->ptr++;
263     return tString;
264 }
265 
266 static int parse_numeric_literal(parser_ctx_t *ctx, void **ret)
267 {
268     BOOL use_int = TRUE;
269     LONGLONG d = 0, hlp;
270     int exp = 0;
271     double r;
272 
273     if(*ctx->ptr == '0' && !('0' <= ctx->ptr[1] && ctx->ptr[1] <= '9') && ctx->ptr[1] != '.')
274         return *ctx->ptr++;
275 
276     while(ctx->ptr < ctx->end && iswdigit(*ctx->ptr)) {
277         hlp = d*10 + *(ctx->ptr++) - '0';
278         if(d>MAXLONGLONG/10 || hlp<0) {
279             exp++;
280             break;
281         }
282         else
283             d = hlp;
284     }
285     while(ctx->ptr < ctx->end && iswdigit(*ctx->ptr)) {
286         exp++;
287         ctx->ptr++;
288     }
289 
290     if(*ctx->ptr == '.') {
291         use_int = FALSE;
292         ctx->ptr++;
293 
294         while(ctx->ptr < ctx->end && iswdigit(*ctx->ptr)) {
295             hlp = d*10 + *(ctx->ptr++) - '0';
296             if(d>MAXLONGLONG/10 || hlp<0)
297                 break;
298 
299             d = hlp;
300             exp--;
301         }
302         while(ctx->ptr < ctx->end && iswdigit(*ctx->ptr))
303             ctx->ptr++;
304     }
305 
306     if(*ctx->ptr == 'e' || *ctx->ptr == 'E') {
307         int e = 0, sign = 1;
308 
309         ctx->ptr++;
310         if(*ctx->ptr == '-') {
311             ctx->ptr++;
312             sign = -1;
313         }else if(*ctx->ptr == '+') {
314             ctx->ptr++;
315         }
316 
317         if(!iswdigit(*ctx->ptr)) {
318             FIXME("Invalid numeric literal\n");
319             return 0;
320         }
321 
322         use_int = FALSE;
323 
324         do {
325             e = e*10 + *(ctx->ptr++) - '0';
326             if(sign == -1 && -e+exp < -(INT_MAX/100)) {
327                 /* The literal will be rounded to 0 anyway. */
328                 while(iswdigit(*ctx->ptr))
329                     ctx->ptr++;
330                 *(double*)ret = 0;
331                 return tDouble;
332             }
333 
334             if(sign*e + exp > INT_MAX/100) {
335                 FIXME("Invalid numeric literal\n");
336                 return 0;
337             }
338         } while(iswdigit(*ctx->ptr));
339 
340         exp += sign*e;
341     }
342 
343     if(use_int && (LONG)d == d) {
344         *(LONG*)ret = d;
345         return tInt;
346     }
347 
348     r = exp>=0 ? d*pow(10, exp) : d/pow(10, -exp);
349     if(isinf(r)) {
350         FIXME("Invalid numeric literal\n");
351         return 0;
352     }
353 
354     *(double*)ret = r;
355     return tDouble;
356 }
357 
358 static int hex_to_int(WCHAR c)
359 {
360     if('0' <= c && c <= '9')
361         return c-'0';
362     if('a' <= c && c <= 'f')
363         return c+10-'a';
364     if('A' <= c && c <= 'F')
365         return c+10-'A';
366     return -1;
367 }
368 
369 static int parse_hex_literal(parser_ctx_t *ctx, LONG *ret)
370 {
371     const WCHAR *begin = ctx->ptr;
372     LONG l = 0, d;
373 
374     while((d = hex_to_int(*++ctx->ptr)) != -1)
375         l = l*16 + d;
376 
377     if(begin + 9 /* max digits+1 */ < ctx->ptr || (*ctx->ptr != '&' && is_identifier_char(*ctx->ptr))) {
378         FIXME("invalid literal\n");
379         return 0;
380     }
381 
382     if(*ctx->ptr == '&')
383         ctx->ptr++;
384 
385     *ret = l;
386     return tInt;
387 }
388 
389 static void skip_spaces(parser_ctx_t *ctx)
390 {
391     while(*ctx->ptr == ' ' || *ctx->ptr == '\t')
392         ctx->ptr++;
393 }
394 
395 static int comment_line(parser_ctx_t *ctx)
396 {
397     static const WCHAR newlineW[] = {'\n','\r',0};
398     ctx->ptr = wcspbrk(ctx->ptr, newlineW);
399     if(ctx->ptr)
400         ctx->ptr++;
401     else
402         ctx->ptr = ctx->end;
403     return tNL;
404 }
405 
406 static int parse_next_token(void *lval, parser_ctx_t *ctx)
407 {
408     WCHAR c;
409 
410     skip_spaces(ctx);
411     if(ctx->ptr == ctx->end)
412         return ctx->last_token == tNL ? tEOF : tNL;
413 
414     c = *ctx->ptr;
415 
416     if('0' <= c && c <= '9')
417         return parse_numeric_literal(ctx, lval);
418 
419     if(iswalpha(c)) {
420         int ret = check_keywords(ctx, lval);
421         if(!ret)
422             return parse_identifier(ctx, lval);
423         if(ret != tREM)
424             return ret;
425         c = '\'';
426     }
427 
428     switch(c) {
429     case '\n':
430     case '\r':
431         ctx->ptr++;
432         return tNL;
433     case '\'':
434         return comment_line(ctx);
435     case ':':
436     case ')':
437     case ',':
438     case '=':
439     case '+':
440     case '*':
441     case '/':
442     case '^':
443     case '\\':
444     case '.':
445     case '_':
446         return *ctx->ptr++;
447     case '-':
448         if(ctx->is_html && ctx->ptr[1] == '-' && ctx->ptr[2] == '>')
449             return comment_line(ctx);
450         ctx->ptr++;
451         return '-';
452     case '(':
453         /* NOTE:
454          * We resolve empty brackets in lexer instead of parser to avoid complex conflicts
455          * in call statement special case |f()| without 'call' keyword
456          */
457         ctx->ptr++;
458         skip_spaces(ctx);
459         if(*ctx->ptr == ')') {
460             ctx->ptr++;
461             return tEMPTYBRACKETS;
462         }
463         return '(';
464     case '"':
465         return parse_string_literal(ctx, lval);
466     case '&':
467         if(*++ctx->ptr == 'h' || *ctx->ptr == 'H')
468             return parse_hex_literal(ctx, lval);
469         return '&';
470     case '<':
471         switch(*++ctx->ptr) {
472         case '>':
473             ctx->ptr++;
474             return tNEQ;
475         case '=':
476             ctx->ptr++;
477             return tLTEQ;
478         case '!':
479             if(ctx->is_html && ctx->ptr[1] == '-' && ctx->ptr[2] == '-')
480                 return comment_line(ctx);
481         }
482         return '<';
483     case '>':
484         if(*++ctx->ptr == '=') {
485             ctx->ptr++;
486             return tGTEQ;
487         }
488         return '>';
489     default:
490         FIXME("Unhandled char %c in %s\n", *ctx->ptr, debugstr_w(ctx->ptr));
491     }
492 
493     return 0;
494 }
495 
496 int parser_lex(void *lval, parser_ctx_t *ctx)
497 {
498     int ret;
499 
500     if (ctx->last_token == tEXPRESSION)
501     {
502         ctx->last_token = tNL;
503         return tEXPRESSION;
504     }
505 
506     while(1) {
507         ret = parse_next_token(lval, ctx);
508         if(ret == '_') {
509             skip_spaces(ctx);
510             if(*ctx->ptr != '\n' && *ctx->ptr != '\r') {
511                 FIXME("'_' not followed by newline\n");
512                 return 0;
513             }
514             if(*ctx->ptr == '\r')
515                 ctx->ptr++;
516             if(*ctx->ptr == '\n')
517                 ctx->ptr++;
518             continue;
519         }
520         if(ret != tNL || ctx->last_token != tNL)
521             break;
522 
523         ctx->last_nl = ctx->ptr-ctx->code;
524     }
525 
526     return (ctx->last_token = ret);
527 }
528