2 * general implementation of scanf used by scanf, sscanf, fscanf,
3 * _cscanf, wscanf, swscanf and fwscanf
5 * Copyright 1996,1998 Marcus Meissner
6 * Copyright 1996 Jukka Iivonen
7 * Copyright 1997,2000, 2003 Uwe Bonnes
8 * Copyright 2000 Jon Griffiths
9 * Copyright 2002 Daniel Gudbjartsson
11 * This library is free software; you can redistribute it and/or
12 * modify it under the terms of the GNU Lesser General Public
13 * License as published by the Free Software Foundation; either
14 * version 2.1 of the License, or (at your option) any later version.
16 * This library is distributed in the hope that it will be useful,
17 * but WITHOUT ANY WARRANTY; without even the implied warranty of
18 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19 * Lesser General Public License for more details.
21 * You should have received a copy of the GNU Lesser General Public
22 * License along with this library; if not, write to the Free Software
23 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
27 #define _CHAR_ wchar_t
29 #define _EOF_RET (short)WEOF
30 #define _ISSPACE_(c) iswspace(c)
31 #define _ISDIGIT_(c) iswdigit(c)
32 #define _WIDE2SUPPORTED_(c) c /* No conversion needed (wide to wide) */
33 #define _CHAR2SUPPORTED_(c) c /* FIXME: convert char to wide char */
34 #define _CHAR2DIGIT_(c, base) wchar2digit((c), (base))
35 #define _BITMAPSIZE_ 256*256
36 #else /* WIDE_SCANF */
40 #define _ISSPACE_(c) isspace((unsigned char)(c))
41 #define _ISDIGIT_(c) isdigit((unsigned char)(c))
42 #define _WIDE2SUPPORTED_(c) c /* FIXME: convert wide char to char */
43 #define _CHAR2SUPPORTED_(c) c /* No conversion needed (char to char) */
44 #define _CHAR2DIGIT_(c, base) char2digit((c), (base))
45 #define _BITMAPSIZE_ 256
46 #endif /* WIDE_SCANF */
49 #define _GETC_(file) (consumed++, _getch())
50 #define _UNGETC_(nch, file) do { _ungetch(nch); consumed--; } while(0)
51 #define _FUNCTION_ int vcscanf(const char *format, va_list ap)
56 #define _GETC_(file) (consumed++, *file++)
57 #define _UNGETC_(nch, file) do { file--; consumed--; } while(0)
59 #define _FUNCTION_ int vswscanf(const wchar_t *file, const wchar_t *format, va_list ap)
60 #else /* WIDE_SCANF */
61 #define _FUNCTION_ int vsscanf(const char *file, const char *format, va_list ap)
62 #endif /* WIDE_SCANF */
65 #define _GETC_(file) (consumed++, fgetwc(file))
66 #define _UNGETC_(nch, file) do { ungetwc(nch, file); consumed--; } while(0)
67 #define _FUNCTION_ int vfwscanf(FILE* file, const wchar_t *format, va_list ap)
68 #else /* WIDE_SCANF */
69 #define _GETC_(file) (consumed++, fgetc(file))
70 #define _UNGETC_(nch, file) do { ungetc(nch, file); consumed--; } while(0)
71 #define _FUNCTION_ int vfscanf(FILE* file, const char *format, va_list ap)
72 #endif /* WIDE_SCANF */
77 int rd
= 0, consumed
= 0;
79 if (!*format
) return 0;
82 TRACE("(%s):\n", debugstr_a(format
));
85 TRACE("%s (%s)\n", file
, debugstr_a(format
));
87 TRACE("%p (%s)\n", file
, debugstr_a(format
));
90 #endif /* WIDE_SCANF */
98 /* a whitespace character in the format string causes scanf to read,
99 * but not store, all consecutive white-space characters in the input
100 * up to the next non-white-space character. One white space character
101 * in the input matches any number (including zero) and combination of
102 * white-space characters in the input. */
103 if (_ISSPACE_(*format
)) {
104 /* skip whitespace */
105 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
108 /* a format specification causes scanf to read and convert characters
109 * in the input into values of a specified type. The value is assigned
110 * to an argument in the argument list. Format specifications have
111 * the form %[*][width][{h | l | I64 | L}]type */
112 else if (*format
== '%') {
113 int st
= 0; int suppress
= 0; int width
= 0;
119 int prefix_finished
= 0;
122 /* look for leading asterisk, which means 'suppress assignment of
128 /* look for width specification */
129 while (_ISDIGIT_(*format
)) {
131 width
+=*format
++ - '0';
133 if (width
==0) width
=-1; /* no width spec seen */
134 /* read prefix (if any) */
135 while (!prefix_finished
) {
137 case 'h': h_prefix
++; break;
139 if(*(format
+1) == 'l') {
145 case 'w': w_prefix
= 1; break;
146 case 'L': L_prefix
= 1; break;
148 if (*(format
+ 1) == '6' &&
149 *(format
+ 2) == '4') {
157 if (!prefix_finished
) format
++;
162 case 'P': /* pointer. */
163 if (sizeof(void *) == sizeof(LONGLONG
)) I64_prefix
= 1;
166 case 'X': /* hexadecimal integer. */
169 case 'o': /* octal integer */
172 case 'u': /* unsigned decimal integer */
175 case 'd': /* signed decimal integer */
178 case 'i': /* generic integer */
181 /* read an integer */
185 /* skip initial whitespace */
186 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
189 if (nch
== '-' || nch
== '+') {
190 negative
= (nch
=='-');
192 if (width
>0) width
--;
194 /* look for leading indication of base */
195 if (width
!=0 && nch
== '0' && *format
!= 'p' && *format
!= 'P') {
197 if (width
>0) width
--;
199 if (width
!=0 && (nch
=='x' || nch
=='X')) {
204 if (width
>0) width
--;
210 /* format %i without indication of base */
213 /* throw away leading zeros */
214 while (width
!=0 && nch
=='0') {
216 if (width
>0) width
--;
219 if (width
!=0 && _CHAR2DIGIT_(nch
, base
)!=-1) {
220 cur
= _CHAR2DIGIT_(nch
, base
);
222 if (width
>0) width
--;
225 /* read until no more digits */
226 while (width
!=0 && (nch
!=_EOF_
) && _CHAR2DIGIT_(nch
, base
)!=-1) {
227 cur
= cur
*base
+ _CHAR2DIGIT_(nch
, base
);
229 if (width
>0) width
--;
233 if (!seendigit
) break; /* not a valid number */
236 #define _SET_NUMBER_(type) *va_arg(ap, type*) = (type)(negative ? -cur : cur)
237 if (I64_prefix
) _SET_NUMBER_(LONGLONG
);
238 else if (l_prefix
) _SET_NUMBER_(LONG
);
239 else if (h_prefix
== 1) _SET_NUMBER_(short int);
240 else _SET_NUMBER_(int);
248 case 'G': { /* read a float */
251 /* skip initial whitespace */
252 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
256 if (nch
== '-' || nch
== '+') {
257 negative
= (nch
=='-');
258 if (width
>0) width
--;
262 /* get first digit. */
264 if (!_ISDIGIT_(nch
)) break;
267 if (width
>0) width
--;
268 /* read until no more digits */
269 while (width
!=0 && (nch
!=_EOF_
) && _ISDIGIT_(nch
)) {
270 cur
= cur
*10 + (nch
- '0');
272 if (width
>0) width
--;
275 cur
= 0; /* MaxPayneDemo Fix: .8 -> 0.8 */
277 /* handle decimals */
278 if (width
!=0 && nch
== '.') {
281 if (width
>0) width
--;
282 while (width
!=0 && (nch
!=_EOF_
) && _ISDIGIT_(nch
)) {
284 cur
+= dec
* (nch
- '0');
286 if (width
>0) width
--;
289 /* handle exponent */
290 if (width
!=0 && (nch
== 'e' || nch
== 'E')) {
291 int exponent
= 0, negexp
= 0;
292 double expcnt
, shift
;
294 if (width
>0) width
--;
295 /* possible sign on the exponent */
296 if (width
!=0 && (nch
=='+' || nch
=='-')) {
299 if (width
>0) width
--;
301 /* exponent digits */
302 while (width
!=0 && (nch
!=_EOF_
) && _ISDIGIT_(nch
)) {
304 exponent
+= (nch
- '0');
306 if (width
>0) width
--;
308 /* update 'cur' with this exponent. */
311 while (exponent
!=0) {
315 expcnt
=expcnt
*expcnt
;
317 cur
= (negexp
? cur
/ shift
: cur
* shift
);
321 if (L_prefix
) _SET_NUMBER_(long double);
322 else if (l_prefix
) _SET_NUMBER_(double);
323 else _SET_NUMBER_(float);
327 /* According to msdn,
328 * 's' reads a character string in a call to fscanf
329 * and 'S' a wide character string and vice versa in a
330 * call to fwscanf. The 'h', 'w' and 'l' prefixes override
331 * this behaviour. 'h' forces reading char * but 'l' and 'w'
332 * force reading WCHAR. */
334 if (w_prefix
|| l_prefix
) goto widecharstring
;
335 else if (h_prefix
) goto charstring
;
337 else goto widecharstring
;
338 #else /* WIDE_SCANF */
339 else goto charstring
;
340 #endif /* WIDE_SCANF */
342 if (w_prefix
|| l_prefix
) goto widecharstring
;
343 else if (h_prefix
) goto charstring
;
345 else goto charstring
;
346 #else /* WIDE_SCANF */
347 else goto widecharstring
;
348 #endif /* WIDE_SCANF */
349 charstring
: { /* read a word into a char */
350 char *sptr
= suppress
? NULL
: va_arg(ap
, char*);
351 /* skip initial whitespace */
352 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
354 /* read until whitespace */
355 while (width
!=0 && (nch
!=_EOF_
) && !_ISSPACE_(nch
)) {
356 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
359 if (width
>0) width
--;
362 if (st
&& !suppress
) *sptr
= 0;
365 widecharstring
: { /* read a word into a wchar_t* */
366 wchar_t *sptr
= suppress
? NULL
: va_arg(ap
, wchar_t*);
367 /* skip initial whitespace */
368 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
370 /* read until whitespace */
371 while (width
!=0 && (nch
!=_EOF_
) && !_ISSPACE_(nch
)) {
372 if (!suppress
) *sptr
++ = _WIDE2SUPPORTED_(nch
);
375 if (width
>0) width
--;
378 if (st
&& !suppress
) *sptr
= 0;
381 /* 'c' and 'C work analogously to 's' and 'S' as described
384 if (w_prefix
|| l_prefix
) goto widecharacter
;
385 else if (h_prefix
) goto character
;
387 else goto widecharacter
;
388 #else /* WIDE_SCANF */
390 #endif /* WIDE_SCANF */
392 if (w_prefix
|| l_prefix
) goto widecharacter
;
393 else if (h_prefix
) goto character
;
396 #else /* WIDE_SCANF */
397 else goto widecharacter
;
398 #endif /* WIDE_SCANF */
399 character
: { /* read single character into char */
400 char *str
= suppress
? NULL
: va_arg(ap
, char*);
401 if (width
== -1) width
= 1;
402 while ((width
!= 0) && (nch
!= _EOF_
))
404 if (!suppress
) *str
++ = _CHAR2SUPPORTED_(nch
);
411 widecharacter
: { /* read single character into a wchar_t */
412 wchar_t *str
= suppress
? NULL
: va_arg(ap
, wchar_t*);
413 if (width
== -1) width
= 1;
414 while ((width
!= 0) && (nch
!= _EOF_
))
416 if (!suppress
) *str
++ = _WIDE2SUPPORTED_(nch
);
425 int*n
= va_arg(ap
, int*);
428 /* This is an odd one: according to the standard,
429 * "Execution of a %n directive does not increment the
430 * assignment count returned at the completion of
431 * execution" even if it wasn't suppressed with the
432 * '*' flag. The Corrigendum to the standard seems
433 * to contradict this (comment out the assignment to
434 * suppress below if you want to implement these
435 * alternate semantics) but the windows program I'm
436 * looking at expects the behavior I've coded here
437 * (which happens to be what glibc does as well).
444 _CHAR_
*str
= suppress
? NULL
: va_arg(ap
, _CHAR_
*);
448 int invert
= 0; /* Set if we are NOT to find the chars */
450 /* Init our bitmap */
452 Mask
= RtlAllocateHeap(RtlGetProcessHeap(), HEAP_ZERO_MEMORY
, _BITMAPSIZE_
/8);
454 Mask
= HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY
, _BITMAPSIZE_
/8);
456 RtlInitializeBitMap(&bitMask
, Mask
, _BITMAPSIZE_
);
458 /* Read the format */
465 RtlSetBits(&bitMask
, ']', 1);
468 while(*format
&& (*format
!= ']')) {
469 /* According to msdn:
470 * "Note that %[a-z] and %[z-a] are interpreted as equivalent to %[abcde...z]." */
471 if((*format
== '-') && (*(format
+ 1) != ']')) {
472 if ((*(format
- 1)) < *(format
+ 1))
473 RtlSetBits(&bitMask
, *(format
- 1) +1 , *(format
+ 1) - *(format
- 1));
475 RtlSetBits(&bitMask
, *(format
+ 1) , *(format
- 1) - *(format
+ 1));
478 RtlSetBits(&bitMask
, *format
, 1);
481 /* read until char is not suitable */
482 while ((width
!= 0) && (nch
!= _EOF_
)) {
484 if(RtlAreBitsSet(&bitMask
, nch
, 1)) {
485 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
489 if(RtlAreBitsClear(&bitMask
, nch
, 1)) {
490 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
496 if (width
>0) width
--;
499 if (!suppress
) *sptr
= 0;
501 RtlFreeHeap(RtlGetProcessHeap(), 0, Mask
);
503 HeapFree(GetProcessHeap(), 0, Mask
);
508 /* From spec: "if a percent sign is followed by a character
509 * that has no meaning as a format-control character, that
510 * character and the following characters are treated as
511 * an ordinary sequence of characters, that is, a sequence
512 * of characters that must match the input. For example,
513 * to specify that a percent-sign character is to be input,
515 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
518 suppress
= 1; /* whoops no field to be read */
519 st
= 1; /* but we got what we expected */
524 if (st
&& !suppress
) rd
++;
527 /* a non-white-space character causes scanf to read, but not store,
528 * a matching non-white-space character. */
530 /* check for character match */
531 if (nch
== *format
) {
541 TRACE("returning %d\n", rd
);
550 #undef _CHAR2SUPPORTED_
551 #undef _WIDE2SUPPORTED_