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_ MSVCRT_wchar_t
28 #define _EOF_ MSVCRT_WEOF
29 #define _EOF_RET MSVCRT_WEOF
30 #define _ISSPACE_(c) MSVCRT_iswspace(c)
31 #define _ISDIGIT_(c) MSVCRT_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 */
38 #define _EOF_ MSVCRT_EOF
39 #define _EOF_RET MSVCRT_EOF
40 #define _ISSPACE_(c) isspace(c)
41 #define _ISDIGIT_(c) isdigit(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_ static int MSVCRT_vcscanf(const char *format, __ms_va_list ap)
56 #define _GETC_(file) (consumed++, *file++)
57 #define _UNGETC_(nch, file) do { file--; consumed--; } while(0)
59 #define _FUNCTION_ static int MSVCRT_vswscanf(const MSVCRT_wchar_t *file, const MSVCRT_wchar_t *format, __ms_va_list ap)
60 #else /* WIDE_SCANF */
61 #define _FUNCTION_ static int MSVCRT_vsscanf(const char *file, const char *format, __ms_va_list ap)
62 #endif /* WIDE_SCANF */
65 #define _GETC_(file) (consumed++, MSVCRT_fgetwc(file))
66 #define _UNGETC_(nch, file) do { MSVCRT_ungetwc(nch, file); consumed--; } while(0)
67 #define _FUNCTION_ static int MSVCRT_vfwscanf(MSVCRT_FILE* file, const MSVCRT_wchar_t *format, __ms_va_list ap)
68 #else /* WIDE_SCANF */
69 #define _GETC_(file) (consumed++, MSVCRT_fgetc(file))
70 #define _UNGETC_(nch, file) do { MSVCRT_ungetc(nch, file); consumed--; } while(0)
71 #define _FUNCTION_ static int MSVCRT_vfscanf(MSVCRT_FILE* file, const char *format, __ms_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 */
92 if (nch
== _EOF_
) return _EOF_RET
;
95 /* a whitespace character in the format string causes scanf to read,
96 * but not store, all consecutive white-space characters in the input
97 * up to the next non-white-space character. One white space character
98 * in the input matches any number (including zero) and combination of
99 * white-space characters in the input. */
100 if (_ISSPACE_(*format
)) {
101 /* skip whitespace */
102 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
105 /* a format specification causes scanf to read and convert characters
106 * in the input into values of a specified type. The value is assigned
107 * to an argument in the argument list. Format specifications have
108 * the form %[*][width][{h | l | I64 | L}]type */
109 else if (*format
== '%') {
110 int st
= 0; int suppress
= 0; int width
= 0;
116 int prefix_finished
= 0;
119 /* look for leading asterisk, which means 'suppress assignment of
125 /* look for width specification */
126 while (_ISDIGIT_(*format
)) {
128 width
+=*format
++ - '0';
130 if (width
==0) width
=-1; /* no width spec seen */
131 /* read prefix (if any) */
132 while (!prefix_finished
) {
134 case 'h': h_prefix
= 1; break;
135 case 'l': l_prefix
= 1; break;
136 case 'w': w_prefix
= 1; break;
137 case 'L': L_prefix
= 1; break;
139 if (*(format
+ 1) == '6' &&
140 *(format
+ 2) == '4') {
148 if (!prefix_finished
) format
++;
153 case 'P': /* pointer. */
155 case 'X': /* hexadecimal integer. */
158 case 'o': /* octal integer */
161 case 'u': /* unsigned decimal integer */
164 case 'd': /* signed decimal integer */
167 case 'i': /* generic integer */
170 /* read an integer */
174 /* skip initial whitespace */
175 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
178 if (nch
== '-' || nch
== '+') {
179 negative
= (nch
=='-');
181 if (width
>0) width
--;
183 /* look for leading indication of base */
184 if (width
!=0 && nch
== '0' && *format
!= 'p' && *format
!= 'P') {
186 if (width
>0) width
--;
188 if (width
!=0 && (nch
=='x' || nch
=='X')) {
193 if (width
>0) width
--;
199 /* format %i without indication of base */
202 /* throw away leading zeros */
203 while (width
!=0 && nch
=='0') {
205 if (width
>0) width
--;
208 if (width
!=0 && _CHAR2DIGIT_(nch
, base
)!=-1) {
209 cur
= _CHAR2DIGIT_(nch
, base
);
211 if (width
>0) width
--;
214 /* read until no more digits */
215 while (width
!=0 && (nch
!=_EOF_
) && _CHAR2DIGIT_(nch
, base
)!=-1) {
216 cur
= cur
*base
+ _CHAR2DIGIT_(nch
, base
);
218 if (width
>0) width
--;
222 if (!seendigit
) break; /* not a valid number */
225 #define _SET_NUMBER_(type) *va_arg(ap, type*) = negative ? -cur : cur
226 if (I64_prefix
) _SET_NUMBER_(LONGLONG
);
227 else if (l_prefix
) _SET_NUMBER_(long int);
228 else if (h_prefix
) _SET_NUMBER_(short int);
229 else _SET_NUMBER_(int);
237 case 'G': { /* read a float */
240 /* skip initial whitespace */
241 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
244 if (nch
== '-' || nch
== '+') {
245 negative
= (nch
=='-');
246 if (width
>0) width
--;
250 /* get first digit. */
252 if (!_ISDIGIT_(nch
)) break;
255 if (width
>0) width
--;
256 /* read until no more digits */
257 while (width
!=0 && (nch
!=_EOF_
) && _ISDIGIT_(nch
)) {
258 cur
= cur
*10 + (nch
- '0');
260 if (width
>0) width
--;
263 cur
= 0; /* MaxPayneDemo Fix: .8 -> 0.8 */
265 /* handle decimals */
266 if (width
!=0 && nch
== '.') {
269 if (width
>0) width
--;
270 while (width
!=0 && (nch
!=_EOF_
) && _ISDIGIT_(nch
)) {
272 cur
+= dec
* (nch
- '0');
274 if (width
>0) width
--;
277 /* handle exponent */
278 if (width
!=0 && (nch
== 'e' || nch
== 'E')) {
279 int exponent
= 0, negexp
= 0;
282 if (width
>0) width
--;
283 /* possible sign on the exponent */
284 if (width
!=0 && (nch
=='+' || nch
=='-')) {
287 if (width
>0) width
--;
289 /* exponent digits */
290 while (width
!=0 && (nch
!=_EOF_
) && _ISDIGIT_(nch
)) {
292 exponent
+= (nch
- '0');
294 if (width
>0) width
--;
296 /* update 'cur' with this exponent. */
297 expcnt
= negexp
? .1 : 10;
298 while (exponent
!=0) {
302 expcnt
=expcnt
*expcnt
;
307 if (L_prefix
) _SET_NUMBER_(long double);
308 else if (l_prefix
) _SET_NUMBER_(double);
309 else _SET_NUMBER_(float);
313 /* According to msdn,
314 * 's' reads a character string in a call to fscanf
315 * and 'S' a wide character string and vice versa in a
316 * call to fwscanf. The 'h', 'w' and 'l' prefixes override
317 * this behaviour. 'h' forces reading char * but 'l' and 'w'
318 * force reading WCHAR. */
320 if (w_prefix
|| l_prefix
) goto widecharstring
;
321 else if (h_prefix
) goto charstring
;
323 else goto widecharstring
;
324 #else /* WIDE_SCANF */
325 else goto charstring
;
326 #endif /* WIDE_SCANF */
328 if (w_prefix
|| l_prefix
) goto widecharstring
;
329 else if (h_prefix
) goto charstring
;
331 else goto charstring
;
332 #else /* WIDE_SCANF */
333 else goto widecharstring
;
334 #endif /* WIDE_SCANF */
335 charstring
: { /* read a word into a char */
336 char *sptr
= suppress
? NULL
: va_arg(ap
, char*);
337 /* skip initial whitespace */
338 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
340 /* read until whitespace */
341 while (width
!=0 && (nch
!=_EOF_
) && !_ISSPACE_(nch
)) {
342 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
345 if (width
>0) width
--;
348 if (!suppress
) *sptr
= 0;
351 widecharstring
: { /* read a word into a wchar_t* */
352 MSVCRT_wchar_t
*sptr
= suppress
? NULL
: va_arg(ap
, MSVCRT_wchar_t
*);
353 /* skip initial whitespace */
354 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
356 /* read until whitespace */
357 while (width
!=0 && (nch
!=_EOF_
) && !_ISSPACE_(nch
)) {
358 if (!suppress
) *sptr
++ = _WIDE2SUPPORTED_(nch
);
361 if (width
>0) width
--;
364 if (!suppress
) *sptr
= 0;
367 /* 'c' and 'C work analogously to 's' and 'S' as described
370 if (w_prefix
|| l_prefix
) goto widecharacter
;
371 else if (h_prefix
) goto character
;
373 else goto widecharacter
;
374 #else /* WIDE_SCANF */
376 #endif /* WIDE_SCANF */
378 if (w_prefix
|| l_prefix
) goto widecharacter
;
379 else if (h_prefix
) goto character
;
382 #else /* WIDE_SCANF */
383 else goto widecharacter
;
384 #endif /* WIDE_SCANF */
385 character
: { /* read single character into char */
386 char *str
= suppress
? NULL
: va_arg(ap
, char*);
387 if (width
== -1) width
= 1;
388 while ((width
!= 0) && (nch
!= _EOF_
))
390 if (!suppress
) *str
++ = _CHAR2SUPPORTED_(nch
);
397 widecharacter
: { /* read single character into a wchar_t */
398 MSVCRT_wchar_t
*str
= suppress
? NULL
: va_arg(ap
, MSVCRT_wchar_t
*);
399 if (width
== -1) width
= 1;
400 while ((width
!= 0) && (nch
!= _EOF_
))
402 if (!suppress
) *str
++ = _WIDE2SUPPORTED_(nch
);
411 int*n
= va_arg(ap
, int*);
414 /* This is an odd one: according to the standard,
415 * "Execution of a %n directive does not increment the
416 * assignment count returned at the completion of
417 * execution" even if it wasn't suppressed with the
418 * '*' flag. The Corrigendum to the standard seems
419 * to contradict this (comment out the assignment to
420 * suppress below if you want to implement these
421 * alternate semantics) but the windows program I'm
422 * looking at expects the behavior I've coded here
423 * (which happens to be what glibc does as well).
430 _CHAR_
*str
= suppress
? NULL
: va_arg(ap
, _CHAR_
*);
434 int invert
= 0; /* Set if we are NOT to find the chars */
436 /* Init our bitmap */
437 Mask
= HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY
, _BITMAPSIZE_
/8);
438 RtlInitializeBitMap(&bitMask
, Mask
, _BITMAPSIZE_
);
440 /* Read the format */
447 RtlSetBits(&bitMask
, ']', 1);
450 while(*format
&& (*format
!= ']')) {
451 /* According to msdn:
452 * "Note that %[a-z] and %[z-a] are interpreted as equivalent to %[abcde...z]." */
453 if((*format
== '-') && (*(format
+ 1) != ']')) {
454 if ((*(format
- 1)) < *(format
+ 1))
455 RtlSetBits(&bitMask
, *(format
- 1) +1 , *(format
+ 1) - *(format
- 1));
457 RtlSetBits(&bitMask
, *(format
+ 1) , *(format
- 1) - *(format
+ 1));
460 RtlSetBits(&bitMask
, *format
, 1);
463 /* read until char is not suitable */
464 while ((width
!= 0) && (nch
!= _EOF_
)) {
466 if(RtlAreBitsSet(&bitMask
, nch
, 1)) {
467 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
471 if(RtlAreBitsClear(&bitMask
, nch
, 1)) {
472 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
478 if (width
>0) width
--;
481 if (!suppress
) *sptr
= 0;
482 HeapFree(GetProcessHeap(), 0, Mask
);
486 /* From spec: "if a percent sign is followed by a character
487 * that has no meaning as a format-control character, that
488 * character and the following characters are treated as
489 * an ordinary sequence of characters, that is, a sequence
490 * of characters that must match the input. For example,
491 * to specify that a percent-sign character is to be input,
493 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
496 suppress
= 1; /* whoops no field to be read */
497 st
= 1; /* but we got what we expected */
502 if (st
&& !suppress
) rd
++;
505 /* a non-white-space character causes scanf to read, but not store,
506 * a matching non-white-space character. */
508 /* check for character match */
509 if (nch
== *format
) {
518 TRACE("returning %d\n", rd
);
527 #undef _CHAR2SUPPORTED_
528 #undef _WIDE2SUPPORTED_