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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 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, 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, va_list ap)
60 #else /* WIDE_SCANF */
61 #define _FUNCTION_ static int MSVCRT_vsscanf(const char *file, const char *format, 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, 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, va_list ap)
72 #endif /* WIDE_SCANF */
76 /*********************************************************************
77 * Implemented based on
78 * http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vccore98/html/_crt_format_specification_fields_.2d_.scanf_and_wscanf_functions.asp
79 * Extended by C. Scott Ananian <cananian@alumni.princeton.edu> to handle
80 * more types of format spec.
83 int rd
= 0, consumed
= 0;
85 if (!*format
) return 0;
88 WARN("(%s): semi-stub\n", debugstr_a(format
));
91 WARN("%s (%s): semi-stub\n", file
, debugstr_a(format
));
93 WARN("%p (%s): semi-stub\n", file
, debugstr_a(format
));
96 #endif /* WIDE_SCANF */
98 if (nch
== _EOF_
) return _EOF_RET
;
101 /* a whitespace character in the format string causes scanf to read,
102 * but not store, all consecutive white-space characters in the input
103 * up to the next non-white-space character. One white space character
104 * in the input matches any number (including zero) and combination of
105 * white-space characters in the input. */
106 if (_ISSPACE_(*format
)) {
107 /* skip whitespace */
108 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
111 /* a format specification causes scanf to read and convert characters
112 * in the input into values of a specified type. The value is assigned
113 * to an argument in the argument list. Format specifications have
114 * the form %[*][width][{h | l | I64 | L}]type */
115 else if (*format
== '%') {
116 int st
= 0; int suppress
= 0; int width
= 0;
117 int base
, number_signed
;
122 int prefix_finished
= 0;
123 /* int I64_prefix = 0; */
125 /* look for leading asterisk, which means 'suppress assignment of
131 /* look for width specification */
132 while (_ISDIGIT_(*format
)) {
134 width
+=*format
++ - '0';
136 if (width
==0) width
=-1; /* no width spec seen */
137 /* read prefix (if any) */
138 while (!prefix_finished
) {
140 case 'h': h_prefix
= 1; break;
141 case 'l': l_prefix
= 1; break;
142 case 'w': w_prefix
= 1; break;
143 case 'L': L_prefix
= 1; break;
145 if (*(format
+ 1) == '6' &&
146 *(format
+ 2) == '4') {
147 /* I64_prefix = 1; */
149 FIXME("I64 prefix currently not implemented in fscanf/fwscanf");
155 if (!prefix_finished
) format
++;
160 case 'X': /* hexadecimal integer. */
161 base
= 16; number_signed
= 0;
163 case 'o': /* octal integer */
164 base
= 8; number_signed
= 0;
166 case 'u': /* unsigned decimal integer */
167 base
= 10; number_signed
= 0;
169 case 'd': /* signed decimal integer */
170 base
= 10; number_signed
= 1;
172 case 'i': /* generic integer */
173 base
= 10; number_signed
= 1;
175 /* read an integer */
176 long unsigned int cur
= 0;
179 /* skip initial whitespace */
180 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
183 if (number_signed
&& (nch
== '-' ||
185 negative
= (nch
=='-');
187 if (width
>0) width
--;
189 /* look for leading indication of base */
190 if (width
!=0 && nch
== '0') {
192 if (width
>0) width
--;
194 if (width
!=0 && (nch
=='x' || nch
=='X')) {
199 if (width
>0) width
--;
205 /* throw away leading zeros */
206 while (width
!=0 && nch
=='0') {
208 if (width
>0) width
--;
211 if (width
!=0 && _CHAR2DIGIT_(nch
, base
)!=-1) {
212 cur
= _CHAR2DIGIT_(nch
, base
);
214 if (width
>0) width
--;
217 /* read until no more digits */
218 while (width
!=0 && (nch
!=_EOF_
) && _CHAR2DIGIT_(nch
, base
)!=-1) {
219 cur
= cur
*base
+ _CHAR2DIGIT_(nch
, base
);
221 if (width
>0) width
--;
225 if (!seendigit
) break; /* not a valid number */
228 #define _SET_NUMBER_(type) *va_arg(ap, type*) = negative ? -cur : cur
230 if (l_prefix
) _SET_NUMBER_(long int);
231 else if (h_prefix
) _SET_NUMBER_(short int);
232 else _SET_NUMBER_(int);
235 WARN("Dropping sign in reading a negative number into an unsigned value");
238 if (l_prefix
) _SET_NUMBER_(unsigned long int);
240 _SET_NUMBER_(unsigned short int);
241 else _SET_NUMBER_(unsigned int);
250 case 'G': { /* read a float */
253 /* skip initial whitespace */
254 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
257 if (nch
== '-' || nch
== '+') {
258 negative
= (nch
=='-');
259 if (width
>0) width
--;
263 /* 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
--;
274 /* handle decimals */
275 if (width
!=0 && nch
== '.') {
278 if (width
>0) width
--;
279 while (width
!=0 && (nch
!=_EOF_
) && _ISDIGIT_(nch
)) {
281 cur
+= dec
* (nch
- '0');
283 if (width
>0) width
--;
286 /* handle exponent */
287 if (width
!=0 && (nch
== 'e' || nch
== 'E')) {
288 int exponent
= 0, negexp
= 0;
291 if (width
>0) width
--;
292 /* possible sign on the exponent */
293 if (width
!=0 && (nch
=='+' || nch
=='-')) {
296 if (width
>0) width
--;
298 /* exponent digits */
299 while (width
!=0 && (nch
!=_EOF_
) && _ISDIGIT_(nch
)) {
301 exponent
+= (nch
- '0');
303 if (width
>0) width
--;
305 /* update 'cur' with this exponent. */
306 expcnt
= negexp
? .1 : 10;
307 while (exponent
!=0) {
311 expcnt
=expcnt
*expcnt
;
316 if (L_prefix
) _SET_NUMBER_(long double);
317 else if (l_prefix
) _SET_NUMBER_(double);
318 else _SET_NUMBER_(float);
323 * http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vclib/html/_crt_scanf_type_field_characters.asp
324 * 's' reads a character string in a call to fscanf
325 * and 'S' a wide character string and vice versa in a
326 * call to fwscanf. The 'h', 'w' and 'l' prefixes override
327 * this behaviour. 'h' forces reading char * but 'l' and 'w'
328 * force reading WCHAR. */
330 if (w_prefix
|| l_prefix
) goto widecharstring
;
331 else if (h_prefix
) goto charstring
;
333 else goto widecharstring
;
334 #else /* WIDE_SCANF */
335 else goto charstring
;
336 #endif /* WIDE_SCANF */
338 if (w_prefix
|| l_prefix
) goto widecharstring
;
339 else if (h_prefix
) goto charstring
;
341 else goto charstring
;
342 #else /* WIDE_SCANF */
343 else goto widecharstring
;
344 #endif /* WIDE_SCANF */
345 charstring
: { /* read a word into a char */
346 char*str
= suppress
? NULL
: va_arg(ap
, char*);
348 /* skip initial whitespace */
349 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
351 /* read until whitespace */
352 while (width
!=0 && (nch
!=_EOF_
) && !_ISSPACE_(nch
)) {
353 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
356 if (width
>0) width
--;
359 if (!suppress
) *sptr
= 0;
362 widecharstring
: { /* read a word into a wchar_t* */
364 suppress
? NULL
: va_arg(ap
, MSVCRT_wchar_t
*);
365 MSVCRT_wchar_t
*sptr
= str
;
366 /* skip initial whitespace */
367 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
369 /* read until whitespace */
370 while (width
!=0 && (nch
!=_EOF_
) && !_ISSPACE_(nch
)) {
371 if (!suppress
) *sptr
++ = _WIDE2SUPPORTED_(nch
);
374 if (width
>0) width
--;
377 if (!suppress
) *sptr
= 0;
380 /* 'c' and 'C work analogously to 's' and 'S' as described
383 if (w_prefix
|| l_prefix
) goto widecharacter
;
384 else if (h_prefix
) goto character
;
386 else goto widecharacter
;
387 #else /* WIDE_SCANF */
389 #endif /* WIDE_SCANF */
391 if (w_prefix
|| l_prefix
) goto widecharacter
;
392 else if (h_prefix
) goto character
;
395 #else /* WIDE_SCANF */
396 else goto widecharacter
;
397 #endif /* WIDE_SCANF */
398 character
: { /* read single character into char */
400 char*c
= va_arg(ap
, char*);
401 *c
= _CHAR2SUPPORTED_(nch
);
408 if (!suppress
) { /* read single character into a wchar_t */
409 MSVCRT_wchar_t
*c
= va_arg(ap
, MSVCRT_wchar_t
*);
410 *c
= _WIDE2SUPPORTED_(nch
);
418 int*n
= va_arg(ap
, int*);
419 *n
= consumed
- (nch
!=_EOF_
);
424 _CHAR_
*str
= suppress
? NULL
: va_arg(ap
, _CHAR_
*);
428 int invert
= 0; /* Set if we are NOT to find the chars */
430 /* Init our bitmap */
431 Mask
= HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY
, _BITMAPSIZE_
/8);
432 RtlInitializeBitMap(&bitMask
, Mask
, _BITMAPSIZE_
);
434 /* Read the format */
441 RtlSetBits(&bitMask
, ']', 1);
444 while(*format
&& (*format
!= ']')) {
446 * http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vccore98/html/_crt_scanf_width_specification.asp
447 * "Note that %[a-z] and %[z-a] are interpreted as equivalent to %[abcde...z]." */
448 if((*format
== '-') && (*(format
+ 1) != ']')) {
449 if ((*(format
- 1)) < *(format
+ 1))
450 RtlSetBits(&bitMask
, *(format
- 1) +1 , *(format
+ 1) - *(format
- 1));
452 RtlSetBits(&bitMask
, *(format
+ 1) , *(format
- 1) - *(format
+ 1));
455 RtlSetBits(&bitMask
, *format
, 1);
458 /* read until char is not suitable */
459 while ((width
!= 0) && (nch
!= _EOF_
)) {
461 if(RtlAreBitsSet(&bitMask
, nch
, 1)) {
462 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
466 if(RtlAreBitsClear(&bitMask
, nch
, 1)) {
467 if (!suppress
) *sptr
++ = _CHAR2SUPPORTED_(nch
);
473 if (width
>0) width
--;
476 if (!suppress
) *sptr
= 0;
477 HeapFree(GetProcessHeap(), 0, Mask
);
481 /* From spec: "if a percent sign is followed by a character
482 * that has no meaning as a format-control character, that
483 * character and the following characters are treated as
484 * an ordinary sequence of characters, that is, a sequence
485 * of characters that must match the input. For example,
486 * to specify that a percent-sign character is to be input,
488 while ((nch
!=_EOF_
) && _ISSPACE_(nch
))
491 suppress
= 1; /* whoops no field to be read */
492 st
= 1; /* but we got what we expected */
497 if (st
&& !suppress
) rd
++;
500 /* a non-white-space character causes scanf to read, but not store,
501 * a matching non-white-space character. */
503 /* check for character match */
504 if (nch
== *format
) {
513 TRACE("returning %d\n", rd
);
522 #undef _CHAR2SUPPORTED_
523 #undef _WIDE2SUPPORTED_