stdio-common/printf-parsemb.c

   1 /* Helper functions for parsing printf format strings.
   2    Copyright (C) 1995-2000,2002,2003,2004,2006 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 /* FORMAT must point to a '%' at the beginning of a spec.  Fills in *SPEC
  50    with the parsed details.  POSN is the number of arguments already
  51    consumed.  At most MAXTYPES - POSN types are filled in TYPES.  Return
  52    the number of args consumed by this spec; *MAX_REF_ARG is updated so it
  53    remains the highest argument index used.  */
  54 size_t
  55 attribute_hidden
  56 #ifdef COMPILE_WPRINTF
  57 __parse_one_specwc (const UCHAR_T *format, size_t posn,
  58                     struct printf_spec *spec, size_t *max_ref_arg)
  59 #else
  60 __parse_one_specmb (const UCHAR_T *format, size_t posn,
  61                     struct printf_spec *spec, size_t *max_ref_arg)
  62 #endif
  63 {
  64   unsigned int n;
  65   size_t nargs = 0;
  66
  67   /* Skip the '%'.  */
  68   ++format;
  69
  70   /* Clear information structure.  */
  71   spec->data_arg = -1;
  72   spec->info.alt = 0;
  73   spec->info.space = 0;
  74   spec->info.left = 0;
  75   spec->info.showsign = 0;
  76   spec->info.group = 0;
  77   spec->info.i18n = 0;
  78   spec->info.extra = 0;
  79   spec->info.pad = ' ';
  80   spec->info.wide = sizeof (UCHAR_T) > 1;
  81
  82   /* Test for positional argument.  */
  83   if (ISDIGIT (*format))
  84     {
  85       const UCHAR_T *begin = format;
  86
  87       n = read_int (&format);
  88
  89       if (n > 0 && *format == L_('$'))
  90         /* Is positional parameter.  */
  91         {
  92           ++format;             /* Skip the '$'.  */
  93           spec->data_arg = n - 1;
  94           *max_ref_arg = MAX (*max_ref_arg, n);
  95         }
  96       else
  97         /* Oops; that was actually the width and/or 0 padding flag.
  98            Step back and read it again.  */
  99         format = begin;
 100     }
 101
 102   /* Check for spec modifiers.  */
 103   do
 104     {
 105       switch (*format)
 106         {
 107         case L_(' '):
 108           /* Output a space in place of a sign, when there is no sign.  */
 109           spec->info.space = 1;
 110           continue;
 111         case L_('+'):
 112           /* Always output + or - for numbers.  */
 113           spec->info.showsign = 1;
 114           continue;
 115         case L_('-'):
 116           /* Left-justify things.  */
 117           spec->info.left = 1;
 118           continue;
 119         case L_('#'):
 120           /* Use the "alternate form":
 121              Hex has 0x or 0X, FP always has a decimal point.  */
 122           spec->info.alt = 1;
 123           continue;
 124         case L_('0'):
 125           /* Pad with 0s.  */
 126           spec->info.pad = '0';
 127           continue;
 128         case L_('\''):
 129           /* Show grouping in numbers if the locale information
 130              indicates any.  */
 131           spec->info.group = 1;
 132           continue;
 133         case L_('I'):
 134           /* Use the internationalized form of the output.  Currently
 135              means to use the `outdigits' of the current locale.  */
 136           spec->info.i18n = 1;
 137           continue;
 138         default:
 139           break;
 140         }
 141       break;
 142     }
 143   while (*++format);
 144
 145   if (spec->info.left)
 146     spec->info.pad = ' ';
 147
 148   /* Get the field width.  */
 149   spec->width_arg = -1;
 150   spec->info.width = 0;
 151   if (*format == L_('*'))
 152     {
 153       /* The field width is given in an argument.
 154          A negative field width indicates left justification.  */
 155       const UCHAR_T *begin = ++format;
 156
 157       if (ISDIGIT (*format))
 158         {
 159           /* The width argument might be found in a positional parameter.  */
 160           n = read_int (&format);
 161
 162           if (n > 0 && *format == L_('$'))
 163             {
 164               spec->width_arg = n - 1;
 165               *max_ref_arg = MAX (*max_ref_arg, n);
 166               ++format;         /* Skip '$'.  */
 167             }
 168         }
 169
 170       if (spec->width_arg < 0)
 171         {
 172           /* Not in a positional parameter.  Consume one argument.  */
 173           spec->width_arg = posn++;
 174           ++nargs;
 175           format = begin;       /* Step back and reread.  */
 176         }
 177     }
 178   else if (ISDIGIT (*format))
 179     /* Constant width specification.  */
 180     spec->info.width = read_int (&format);
 181
 182   /* Get the precision.  */
 183   spec->prec_arg = -1;
 184   /* -1 means none given; 0 means explicit 0.  */
 185   spec->info.prec = -1;
 186   if (*format == L_('.'))
 187     {
 188       ++format;
 189       if (*format == L_('*'))
 190         {
 191           /* The precision is given in an argument.  */
 192           const UCHAR_T *begin = ++format;
 193
 194           if (ISDIGIT (*format))
 195             {
 196               n = read_int (&format);
 197
 198               if (n > 0 && *format == L_('$'))
 199                 {
 200                   spec->prec_arg = n - 1;
 201                   *max_ref_arg = MAX (*max_ref_arg, n);
 202                   ++format;
 203                 }
 204             }
 205
 206           if (spec->prec_arg < 0)
 207             {
 208               /* Not in a positional parameter.  */
 209               spec->prec_arg = posn++;
 210               ++nargs;
 211               format = begin;
 212             }
 213         }
 214       else if (ISDIGIT (*format))
 215         spec->info.prec = read_int (&format);
 216       else
 217         /* "%.?" is treated like "%.0?".  */
 218         spec->info.prec = 0;
 219     }
 220
 221   /* Check for type modifiers.  */
 222   spec->info.is_long_double = 0;
 223   spec->info.is_short = 0;
 224   spec->info.is_long = 0;
 225   spec->info.is_char = 0;
 226
 227   switch (*format++)
 228     {
 229     case L_('h'):
 230       /* ints are short ints or chars.  */
 231       if (*format != L_('h'))
 232         spec->info.is_short = 1;
 233       else
 234         {
 235           ++format;
 236           spec->info.is_char = 1;
 237         }
 238       break;
 239     case L_('l'):
 240       /* ints are long ints.  */
 241       spec->info.is_long = 1;
 242       if (*format != L_('l'))
 243         break;
 244       ++format;
 245       /* FALLTHROUGH */
 246     case L_('L'):
 247       /* doubles are long doubles, and ints are long long ints.  */
 248     case L_('q'):
 249       /* 4.4 uses this for long long.  */
 250       spec->info.is_long_double = 1;
 251       break;
 252     case L_('z'):
 253     case L_('Z'):
 254       /* ints are size_ts.  */
 255       assert (sizeof (size_t) <= sizeof (unsigned long long int));
 256 #if LONG_MAX != LONG_LONG_MAX
 257       spec->info.is_long_double = sizeof (size_t) > sizeof (unsigned long int);
 258 #endif
 259       spec->info.is_long = sizeof (size_t) > sizeof (unsigned int);
 260       break;
 261     case L_('t'):
 262       assert (sizeof (ptrdiff_t) <= sizeof (long long int));
 263 #if LONG_MAX != LONG_LONG_MAX
 264       spec->info.is_long_double = (sizeof (ptrdiff_t) > sizeof (long int));
 265 #endif
 266       spec->info.is_long = sizeof (ptrdiff_t) > sizeof (int);
 267       break;
 268     case L_('j'):
 269       assert (sizeof (uintmax_t) <= sizeof (unsigned long long int));
 270 #if LONG_MAX != LONG_LONG_MAX
 271       spec->info.is_long_double = (sizeof (uintmax_t)
 272                                    > sizeof (unsigned long int));
 273 #endif
 274       spec->info.is_long = sizeof (uintmax_t) > sizeof (unsigned int);
 275       break;
 276     default:
 277       /* Not a recognized modifier.  Backup.  */
 278       --format;
 279       break;
 280     }
 281
 282   /* Get the format specification.  */
 283   spec->info.spec = (wchar_t) *format++;
 284   if (__builtin_expect (__printf_function_table != NULL, 0)
 285       && spec->info.spec <= UCHAR_MAX
 286       && __printf_arginfo_table[spec->info.spec] != NULL)
 287     /* We don't try to get the types for all arguments if the format
 288        uses more than one.  The normal case is covered though.  */
 289     spec->ndata_args = (*__printf_arginfo_table[spec->info.spec])
 290       (&spec->info, 1, &spec->data_arg_type);
 291   else
 292     {
 293       /* Find the data argument types of a built-in spec.  */
 294       spec->ndata_args = 1;
 295
 296       switch (spec->info.spec)
 297         {
 298         case L'i':
 299         case L'd':
 300         case L'u':
 301         case L'o':
 302         case L'X':
 303         case L'x':
 304 #if LONG_MAX != LONG_LONG_MAX
 305           if (spec->info.is_long_double)
 306             spec->data_arg_type = PA_INT|PA_FLAG_LONG_LONG;
 307           else
 308 #endif
 309             if (spec->info.is_long)
 310               spec->data_arg_type = PA_INT|PA_FLAG_LONG;
 311             else if (spec->info.is_short)
 312               spec->data_arg_type = PA_INT|PA_FLAG_SHORT;
 313             else if (spec->info.is_char)
 314               spec->data_arg_type = PA_CHAR;
 315             else
 316               spec->data_arg_type = PA_INT;
 317           break;
 318         case L'e':
 319         case L'E':
 320         case L'f':
 321         case L'F':
 322         case L'g':
 323         case L'G':
 324         case L'a':
 325         case L'A':
 326           if (spec->info.is_long_double)
 327             spec->data_arg_type = PA_DOUBLE|PA_FLAG_LONG_DOUBLE;
 328           else
 329             spec->data_arg_type = PA_DOUBLE;
 330           break;
 331         case L'c':
 332           spec->data_arg_type = PA_CHAR;
 333           break;
 334         case L'C':
 335           spec->data_arg_type = PA_WCHAR;
 336           break;
 337         case L's':
 338           spec->data_arg_type = PA_STRING;
 339           break;
 340         case L'S':
 341           spec->data_arg_type = PA_WSTRING;
 342           break;
 343         case L'p':
 344           spec->data_arg_type = PA_POINTER;
 345           break;
 346         case L'n':
 347           spec->data_arg_type = PA_INT|PA_FLAG_PTR;
 348           break;
 349
 350         case L'm':
 351         default:
 352           /* An unknown spec will consume no args.  */
 353           spec->ndata_args = 0;
 354           break;
 355         }
 356     }
 357
 358   if (spec->data_arg == -1 && spec->ndata_args > 0)
 359     {
 360       /* There are args consumed, but no positional spec.  Use the
 361          next sequential arg position.  */
 362       spec->data_arg = posn;
 363       nargs += spec->ndata_args;
 364     }
 365
 366   if (spec->info.spec == L'\0')
 367     /* Format ended before this spec was complete.  */
 368     spec->end_of_fmt = spec->next_fmt = format - 1;
 369   else
 370     {
 371       /* Find the next format spec.  */
 372       spec->end_of_fmt = format;
 373 #ifdef COMPILE_WPRINTF
 374       spec->next_fmt = __find_specwc (format);
 375 #else
 376       spec->next_fmt = __find_specmb (format);
 377 #endif
 378     }
 379
 380   return nargs;
 381 }