stdio-common/printf-parsemb.c

   1 /* Helper functions for parsing printf format strings.
   2    Copyright (C) 1995-2000, 2002, 2003, 2004 Free Software Foundation, Inc.
   3    This file is part of th GNU C Library.
   4
   5    The GNU C Library is free software; you can redistribute it and/or
   6    modify it under the terms of the GNU Lesser General Public
   7    License as published by the Free Software Foundation; either
   8    version 2.1 of the License, or (at your option) any later version.
   9
  10    The GNU C Library is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13    Lesser General Public License for more details.
  14
  15    You should have received a copy of the GNU Lesser General Public
  16    License along with the GNU C Library; if not, write to the Free
  17    Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
  18    02111-1307 USA.  */
  19
  20 #include <ctype.h>
  21 #include <limits.h>
  22 #include <stdlib.h>
  23 #include <string.h>
  24 #include <sys/param.h>
  25 #include <wchar.h>
  26 #include <wctype.h>
  27
  28 #ifndef COMPILE_WPRINTF
  29 # define CHAR_T         char
  30 # define UCHAR_T        unsigned char
  31 # define INT_T          int
  32 # define L_(Str)        Str
  33 # define ISDIGIT(Ch)    isdigit (Ch)
  34 #else
  35 # define CHAR_T         wchar_t
  36 # define UCHAR_T        unsigned int
  37 # define INT_T          wint_t
  38 # define L_(Str)        L##Str
  39 # define ISDIGIT(Ch)    iswdigit (Ch)
  40 #endif
  41
  42 #include "printf-parse.h"
  43
  44 #define NDEBUG 1
  45 #include <assert.h>
  46
  47
  48
  49 /* Find the next spec in FORMAT, or the end of the string.  Returns
  50    a pointer into FORMAT, to a '%' or a '\0'.  */
  51 const UCHAR_T *
  52 #ifdef COMPILE_WPRINTF
  53 __find_specwc (const UCHAR_T *format)
  54 #else
  55 __find_specmb (const UCHAR_T *format, mbstate_t *ps)
  56 #endif
  57 {
  58 #ifdef COMPILE_WPRINTF
  59   return (const UCHAR_T *) __wcschrnul ((const CHAR_T *) format, L'%');
  60 #else
  61   while (*format != L_('\0') && *format != L_('%'))
  62     {
  63       int len;
  64
  65       /* Remove any hints of a wrong encoding.  */
  66       ps->__count = 0;
  67       if (! isascii (*format) && (len = __mbrlen (format, MB_CUR_MAX, ps)) > 0)
  68         format += len;
  69       else
  70         ++format;
  71     }
  72   return format;
  73 #endif
  74 }
  75
  76
  77 /* FORMAT must point to a '%' at the beginning of a spec.  Fills in *SPEC
  78    with the parsed details.  POSN is the number of arguments already
  79    consumed.  At most MAXTYPES - POSN types are filled in TYPES.  Return
  80    the number of args consumed by this spec; *MAX_REF_ARG is updated so it
  81    remains the highest argument index used.  */
  82 size_t
  83 attribute_hidden
  84 #ifdef COMPILE_WPRINTF
  85 __parse_one_specwc (const UCHAR_T *format, size_t posn,
  86                     struct printf_spec *spec, size_t *max_ref_arg)
  87 #else
  88 __parse_one_specmb (const UCHAR_T *format, size_t posn,
  89                     struct printf_spec *spec, size_t *max_ref_arg,
  90                     mbstate_t *ps)
  91 #endif
  92 {
  93   unsigned int n;
  94   size_t nargs = 0;
  95
  96   /* Skip the '%'.  */
  97   ++format;
  98
  99   /* Clear information structure.  */
 100   spec->data_arg = -1;
 101   spec->info.alt = 0;
 102   spec->info.space = 0;
 103   spec->info.left = 0;
 104   spec->info.showsign = 0;
 105   spec->info.group = 0;
 106   spec->info.i18n = 0;
 107   spec->info.extra = 0;
 108   spec->info.pad = ' ';
 109   spec->info.wide = sizeof (UCHAR_T) > 1;
 110
 111   /* Test for positional argument.  */
 112   if (ISDIGIT (*format))
 113     {
 114       const UCHAR_T *begin = format;
 115
 116       n = read_int (&format);
 117
 118       if (n > 0 && *format == L_('$'))
 119         /* Is positional parameter.  */
 120         {
 121           ++format;             /* Skip the '$'.  */
 122           spec->data_arg = n - 1;
 123           *max_ref_arg = MAX (*max_ref_arg, n);
 124         }
 125       else
 126         /* Oops; that was actually the width and/or 0 padding flag.
 127            Step back and read it again.  */
 128         format = begin;
 129     }
 130
 131   /* Check for spec modifiers.  */
 132   do
 133     {
 134       switch (*format)
 135         {
 136         case L_(' '):
 137           /* Output a space in place of a sign, when there is no sign.  */
 138           spec->info.space = 1;
 139           continue;
 140         case L_('+'):
 141           /* Always output + or - for numbers.  */
 142           spec->info.showsign = 1;
 143           continue;
 144         case L_('-'):
 145           /* Left-justify things.  */
 146           spec->info.left = 1;
 147           continue;
 148         case L_('#'):
 149           /* Use the "alternate form":
 150              Hex has 0x or 0X, FP always has a decimal point.  */
 151           spec->info.alt = 1;
 152           continue;
 153         case L_('0'):
 154           /* Pad with 0s.  */
 155           spec->info.pad = '0';
 156           continue;
 157         case L_('\''):
 158           /* Show grouping in numbers if the locale information
 159              indicates any.  */
 160           spec->info.group = 1;
 161           continue;
 162         case L_('I'):
 163           /* Use the internationalized form of the output.  Currently
 164              means to use the `outdigits' of the current locale.  */
 165           spec->info.i18n = 1;
 166           continue;
 167         default:
 168           break;
 169         }
 170       break;
 171     }
 172   while (*++format);
 173
 174   if (spec->info.left)
 175     spec->info.pad = ' ';
 176
 177   /* Get the field width.  */
 178   spec->width_arg = -1;
 179   spec->info.width = 0;
 180   if (*format == L_('*'))
 181     {
 182       /* The field width is given in an argument.
 183          A negative field width indicates left justification.  */
 184       const UCHAR_T *begin = ++format;
 185
 186       if (ISDIGIT (*format))
 187         {
 188           /* The width argument might be found in a positional parameter.  */
 189           n = read_int (&format);
 190
 191           if (n > 0 && *format == L_('$'))
 192             {
 193               spec->width_arg = n - 1;
 194               *max_ref_arg = MAX (*max_ref_arg, n);
 195               ++format;         /* Skip '$'.  */
 196             }
 197         }
 198
 199       if (spec->width_arg < 0)
 200         {
 201           /* Not in a positional parameter.  Consume one argument.  */
 202           spec->width_arg = posn++;
 203           ++nargs;
 204           format = begin;       /* Step back and reread.  */
 205         }
 206     }
 207   else if (ISDIGIT (*format))
 208     /* Constant width specification.  */
 209     spec->info.width = read_int (&format);
 210
 211   /* Get the precision.  */
 212   spec->prec_arg = -1;
 213   /* -1 means none given; 0 means explicit 0.  */
 214   spec->info.prec = -1;
 215   if (*format == L_('.'))
 216     {
 217       ++format;
 218       if (*format == L_('*'))
 219         {
 220           /* The precision is given in an argument.  */
 221           const UCHAR_T *begin = ++format;
 222
 223           if (ISDIGIT (*format))
 224             {
 225               n = read_int (&format);
 226
 227               if (n > 0 && *format == L_('$'))
 228                 {
 229                   spec->prec_arg = n - 1;
 230                   *max_ref_arg = MAX (*max_ref_arg, n);
 231                   ++format;
 232                 }
 233             }
 234
 235           if (spec->prec_arg < 0)
 236             {
 237               /* Not in a positional parameter.  */
 238               spec->prec_arg = posn++;
 239               ++nargs;
 240               format = begin;
 241             }
 242         }
 243       else if (ISDIGIT (*format))
 244         spec->info.prec = read_int (&format);
 245       else
 246         /* "%.?" is treated like "%.0?".  */
 247         spec->info.prec = 0;
 248     }
 249
 250   /* Check for type modifiers.  */
 251   spec->info.is_long_double = 0;
 252   spec->info.is_short = 0;
 253   spec->info.is_long = 0;
 254   spec->info.is_char = 0;
 255
 256   switch (*format++)
 257     {
 258     case L_('h'):
 259       /* ints are short ints or chars.  */
 260       if (*format != L_('h'))
 261         spec->info.is_short = 1;
 262       else
 263         {
 264           ++format;
 265           spec->info.is_char = 1;
 266         }
 267       break;
 268     case L_('l'):
 269       /* ints are long ints.  */
 270       spec->info.is_long = 1;
 271       if (*format != L_('l'))
 272         break;
 273       ++format;
 274       /* FALLTHROUGH */
 275     case L_('L'):
 276       /* doubles are long doubles, and ints are long long ints.  */
 277     case L_('q'):
 278       /* 4.4 uses this for long long.  */
 279       spec->info.is_long_double = 1;
 280       break;
 281     case L_('z'):
 282     case L_('Z'):
 283       /* ints are size_ts.  */
 284       assert (sizeof (size_t) <= sizeof (unsigned long long int));
 285 #if LONG_MAX != LONG_LONG_MAX
 286       spec->info.is_long_double = sizeof (size_t) > sizeof (unsigned long int);
 287 #endif
 288       spec->info.is_long = sizeof (size_t) > sizeof (unsigned int);
 289       break;
 290     case L_('t'):
 291       assert (sizeof (ptrdiff_t) <= sizeof (long long int));
 292 #if LONG_MAX != LONG_LONG_MAX
 293       spec->info.is_long_double = (sizeof (ptrdiff_t) > sizeof (long int));
 294 #endif
 295       spec->info.is_long = sizeof (ptrdiff_t) > sizeof (int);
 296       break;
 297     case L_('j'):
 298       assert (sizeof (uintmax_t) <= sizeof (unsigned long long int));
 299 #if LONG_MAX != LONG_LONG_MAX
 300       spec->info.is_long_double = (sizeof (uintmax_t)
 301                                    > sizeof (unsigned long int));
 302 #endif
 303       spec->info.is_long = sizeof (uintmax_t) > sizeof (unsigned int);
 304       break;
 305     default:
 306       /* Not a recognized modifier.  Backup.  */
 307       --format;
 308       break;
 309     }
 310
 311   /* Get the format specification.  */
 312   spec->info.spec = (wchar_t) *format++;
 313   if (__builtin_expect (__printf_function_table != NULL, 0)
 314       && spec->info.spec <= UCHAR_MAX
 315       && __printf_arginfo_table[spec->info.spec] != NULL)
 316     /* We don't try to get the types for all arguments if the format
 317        uses more than one.  The normal case is covered though.  */
 318     spec->ndata_args = (*__printf_arginfo_table[spec->info.spec])
 319       (&spec->info, 1, &spec->data_arg_type);
 320   else
 321     {
 322       /* Find the data argument types of a built-in spec.  */
 323       spec->ndata_args = 1;
 324
 325       switch (spec->info.spec)
 326         {
 327         case L'i':
 328         case L'd':
 329         case L'u':
 330         case L'o':
 331         case L'X':
 332         case L'x':
 333 #if LONG_MAX != LONG_LONG_MAX
 334           if (spec->info.is_long_double)
 335             spec->data_arg_type = PA_INT|PA_FLAG_LONG_LONG;
 336           else
 337 #endif
 338             if (spec->info.is_long)
 339               spec->data_arg_type = PA_INT|PA_FLAG_LONG;
 340             else if (spec->info.is_short)
 341               spec->data_arg_type = PA_INT|PA_FLAG_SHORT;
 342             else if (spec->info.is_char)
 343               spec->data_arg_type = PA_CHAR;
 344             else
 345               spec->data_arg_type = PA_INT;
 346           break;
 347         case L'e':
 348         case L'E':
 349         case L'f':
 350         case L'F':
 351         case L'g':
 352         case L'G':
 353         case L'a':
 354         case L'A':
 355           if (spec->info.is_long_double)
 356             spec->data_arg_type = PA_DOUBLE|PA_FLAG_LONG_DOUBLE;
 357           else
 358             spec->data_arg_type = PA_DOUBLE;
 359           break;
 360         case L'c':
 361           spec->data_arg_type = PA_CHAR;
 362           break;
 363         case L'C':
 364           spec->data_arg_type = PA_WCHAR;
 365           break;
 366         case L's':
 367           spec->data_arg_type = PA_STRING;
 368           break;
 369         case L'S':
 370           spec->data_arg_type = PA_WSTRING;
 371           break;
 372         case L'p':
 373           spec->data_arg_type = PA_POINTER;
 374           break;
 375         case L'n':
 376           spec->data_arg_type = PA_INT|PA_FLAG_PTR;
 377           break;
 378
 379         case L'm':
 380         default:
 381           /* An unknown spec will consume no args.  */
 382           spec->ndata_args = 0;
 383           break;
 384         }
 385     }
 386
 387   if (spec->data_arg == -1 && spec->ndata_args > 0)
 388     {
 389       /* There are args consumed, but no positional spec.  Use the
 390          next sequential arg position.  */
 391       spec->data_arg = posn;
 392       nargs += spec->ndata_args;
 393     }
 394
 395   if (spec->info.spec == L'\0')
 396     /* Format ended before this spec was complete.  */
 397     spec->end_of_fmt = spec->next_fmt = format - 1;
 398   else
 399     {
 400       /* Find the next format spec.  */
 401       spec->end_of_fmt = format;
 402 #ifdef COMPILE_WPRINTF
 403       spec->next_fmt = __find_specwc (format);
 404 #else
 405       spec->next_fmt = __find_specmb (format, ps);
 406 #endif
 407     }
 408
 409   return nargs;
 410 }