stdio-common/printf-parse.h

   1 /* Internal header for parsing printf format strings.
   2    Copyright (C) 1995, 1996 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 Library General Public License as
   7    published by the Free Software Foundation; either version 2 of the
   8    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    Library General Public License for more details.
  14
  15    You should have received a copy of the GNU Library General Public
  16    License along with the GNU C Library; see the file COPYING.LIB.  If not,
  17    write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  18    Boston, MA 02111-1307, USA.  */
  19
  20 #include <ctype.h>
  21 #include <printf.h>
  22 #include <string.h>
  23 #include <stddef.h>
  24
  25 #define NDEBUG 1
  26 #include <assert.h>
  27
  28 #define MAX(a,b)        ({typeof(a) _a = (a); typeof(b) _b = (b);             \
  29                           _a > _b ? _a : _b; })
  30 #define MIN(a,b)        ({typeof(a) _a = (a); typeof(b) _b = (b);             \
  31                           _a < _b ? _a : _b; })
  32
  33 struct printf_spec
  34   {
  35     /* Information parsed from the format spec.  */
  36     struct printf_info info;
  37
  38     /* Pointers into the format string for the end of this format
  39        spec and the next (or to the end of the string if no more).  */
  40     const char *end_of_fmt, *next_fmt;
  41
  42     /* Position of arguments for precision and width, or -1 if `info' has
  43        the constant value.  */
  44     int prec_arg, width_arg;
  45
  46     int data_arg;               /* Position of data argument.  */
  47     int data_arg_type;          /* Type of first argument.  */
  48     /* Number of arguments consumed by this format specifier.  */
  49     size_t ndata_args;
  50   };
  51
  52
  53 /* The various kinds off arguments that can be passed to printf.  */
  54 union printf_arg
  55   {
  56     unsigned char pa_char;
  57     wchar_t pa_wchar;
  58     short int pa_short_int;
  59     int pa_int;
  60     long int pa_long_int;
  61     long long int pa_long_long_int;
  62     unsigned short int pa_u_short_int;
  63     unsigned int pa_u_int;
  64     unsigned long int pa_u_long_int;
  65     unsigned long long int pa_u_long_long_int;
  66     float pa_float;
  67     double pa_double;
  68     long double pa_long_double;
  69     const char *pa_string;
  70     const wchar_t *pa_wstring;
  71     void *pa_pointer;
  72   };
  73
  74
  75 /* Read a simple integer from a string and update the string pointer.
  76    It is assumed that the first character is a digit.  */
  77 static inline unsigned int
  78 read_int (const UCHAR_T * *pstr)
  79 {
  80   unsigned int retval = **pstr - L_('0');
  81
  82   while (ISDIGIT (*++(*pstr)))
  83     {
  84       retval *= 10;
  85       retval += **pstr - L_('0');
  86     }
  87
  88   return retval;
  89 }
  90
  91
  92
  93 /* Find the next spec in FORMAT, or the end of the string.  Returns
  94    a pointer into FORMAT, to a '%' or a '\0'.  */
  95 static inline const char *
  96 find_spec (const char *format, mbstate_t *ps)
  97 {
  98   while (*format != '\0' && *format != '%')
  99     {
 100       int len;
 101
 102       /* Remove any hints of a wrong encoding.  */
 103       ps->count = 0;
 104       if (isascii (*format) || (len = mbrlen (format, MB_CUR_MAX, ps)) <= 0)
 105         ++format;
 106       else
 107         format += len;
 108     }
 109   return format;
 110 }
 111
 112
 113 /* This is defined in reg-printf.c.  */
 114 extern printf_arginfo_function **__printf_arginfo_table;
 115
 116
 117 /* FORMAT must point to a '%' at the beginning of a spec.  Fills in *SPEC
 118    with the parsed details.  POSN is the number of arguments already
 119    consumed.  At most MAXTYPES - POSN types are filled in TYPES.  Return
 120    the number of args consumed by this spec; *MAX_REF_ARG is updated so it
 121    remains the highest argument index used.  */
 122 static inline size_t
 123 parse_one_spec (const UCHAR_T *format, size_t posn, struct printf_spec *spec,
 124                 size_t *max_ref_arg, mbstate_t *ps)
 125 {
 126   unsigned int n;
 127   size_t nargs = 0;
 128
 129   /* Skip the '%'.  */
 130   ++format;
 131
 132   /* Clear information structure.  */
 133   spec->data_arg = -1;
 134   spec->info.alt = 0;
 135   spec->info.space = 0;
 136   spec->info.left = 0;
 137   spec->info.showsign = 0;
 138   spec->info.group = 0;
 139   spec->info.pad = ' ';
 140
 141   /* Test for positional argument.  */
 142   if (ISDIGIT (*format))
 143     {
 144       const UCHAR_T *begin = format;
 145
 146       n = read_int (&format);
 147
 148       if (n > 0 && *format == L_('$'))
 149         /* Is positional parameter.  */
 150         {
 151           ++format;             /* Skip the '$'.  */
 152           spec->data_arg = n - 1;
 153           *max_ref_arg = MAX (*max_ref_arg, n);
 154         }
 155       else
 156         /* Oops; that was actually the width and/or 0 padding flag.
 157            Step back and read it again.  */
 158         format = begin;
 159     }
 160
 161   /* Check for spec modifiers.  */
 162   while (*format == L_(' ') || *format == L_('+') || *format == L_('-') ||
 163          *format == L_('#') || *format == L_('0') || *format == L_('\''))
 164     switch (*format++)
 165       {
 166       case L_(' '):
 167         /* Output a space in place of a sign, when there is no sign.  */
 168         spec->info.space = 1;
 169         break;
 170       case L_('+'):
 171         /* Always output + or - for numbers.  */
 172         spec->info.showsign = 1;
 173         break;
 174       case L_('-'):
 175         /* Left-justify things.  */
 176         spec->info.left = 1;
 177         break;
 178       case L_('#'):
 179         /* Use the "alternate form":
 180            Hex has 0x or 0X, FP always has a decimal point.  */
 181         spec->info.alt = 1;
 182         break;
 183       case L_('0'):
 184         /* Pad with 0s.  */
 185         spec->info.pad = '0';
 186         break;
 187       case L_('\''):
 188         /* Show grouping in numbers if the locale information
 189            indicates any.  */
 190         spec->info.group = 1;
 191         break;
 192       }
 193   if (spec->info.left)
 194     spec->info.pad = ' ';
 195
 196   /* Get the field width.  */
 197   spec->width_arg = -1;
 198   spec->info.width = 0;
 199   if (*format == L_('*'))
 200     {
 201       /* The field width is given in an argument.
 202          A negative field width indicates left justification.  */
 203       const UCHAR_T *begin = ++format;
 204
 205       if (ISDIGIT (*format))
 206         {
 207           /* The width argument might be found in a positional parameter.  */
 208           n = read_int (&format);
 209
 210           if (n > 0 && *format == L_('$'))
 211             {
 212               spec->width_arg = n - 1;
 213               *max_ref_arg = MAX (*max_ref_arg, n);
 214               ++format;         /* Skip '$'.  */
 215             }
 216         }
 217
 218       if (spec->width_arg < 0)
 219         {
 220           /* Not in a positional parameter.  Consume one argument.  */
 221           spec->width_arg = posn++;
 222           ++nargs;
 223           format = begin;       /* Step back and reread.  */
 224         }
 225     }
 226   else if (ISDIGIT (*format))
 227     /* Constant width specification.  */
 228     spec->info.width = read_int (&format);
 229
 230   /* Get the precision.  */
 231   spec->prec_arg = -1;
 232   /* -1 means none given; 0 means explicit 0.  */
 233   spec->info.prec = -1;
 234   if (*format == L_('.'))
 235     {
 236       ++format;
 237       if (*format == L_('*'))
 238         {
 239           /* The precision is given in an argument.  */
 240           const UCHAR_T *begin = ++format;
 241
 242           if (ISDIGIT (*format))
 243             {
 244               n = read_int (&format);
 245
 246               if (n > 0 && *format == L_('$'))
 247                 {
 248                   spec->prec_arg = n - 1;
 249                   *max_ref_arg = MAX (*max_ref_arg, n);
 250                   ++format;
 251                 }
 252             }
 253
 254           if (spec->prec_arg < 0)
 255             {
 256               /* Not in a positional parameter.  */
 257               spec->prec_arg = posn++;
 258               ++nargs;
 259               format = begin;
 260             }
 261         }
 262       else if (ISDIGIT (*format))
 263         spec->info.prec = read_int (&format);
 264       else
 265         /* "%.?" is treated like "%.0?".  */
 266         spec->info.prec = 0;
 267     }
 268
 269   /* Check for type modifiers.  */
 270 #define is_longlong is_long_double
 271   spec->info.is_long_double = 0;
 272   spec->info.is_short = 0;
 273   spec->info.is_long = 0;
 274
 275   if (*format == L_('h') || *format == L_('l') || *format == L_('L') ||
 276       *format == L_('Z') || *format == L_('q'))
 277     switch (*format++)
 278       {
 279       case L_('h'):
 280         /* int's are short int's.  */
 281         spec->info.is_short = 1;
 282         break;
 283       case L_('l'):
 284         /* int's are long int's.  */
 285         spec->info.is_long = 1;
 286         if (*format != L_('l'))
 287           break;
 288         ++format;
 289         /* FALLTHROUGH */
 290       case L_('L'):
 291         /* double's are long double's, and int's are long long int's.  */
 292       case L_('q'):
 293         /* 4.4 uses this for long long.  */
 294         spec->info.is_long_double = 1;
 295         break;
 296       case L_('Z'):
 297         /* int's are size_t's.  */
 298         assert (sizeof(size_t) <= sizeof(unsigned long long int));
 299         spec->info.is_longlong = sizeof(size_t) > sizeof(unsigned long int);
 300         spec->info.is_long = sizeof(size_t) > sizeof(unsigned int);
 301         break;
 302       }
 303
 304   /* Get the format specification.  */
 305   spec->info.spec = (wchar_t) *format++;
 306   if (__printf_arginfo_table != NULL &&
 307       __printf_arginfo_table[spec->info.spec] != NULL)
 308     /* We don't try to get the types for all arguments if the format
 309        uses more than one.  The normal case is covered though.  */
 310     spec->ndata_args = (*__printf_arginfo_table[spec->info.spec])
 311       (&spec->info, 1, &spec->data_arg_type);
 312   else
 313     {
 314       /* Find the data argument types of a built-in spec.  */
 315       spec->ndata_args = 1;
 316
 317       switch (spec->info.spec)
 318         {
 319         case L'i':
 320         case L'd':
 321         case L'u':
 322         case L'o':
 323         case L'X':
 324         case L'x':
 325           if (spec->info.is_longlong)
 326             spec->data_arg_type = PA_INT|PA_FLAG_LONG_LONG;
 327           else if (spec->info.is_long)
 328             spec->data_arg_type = PA_INT|PA_FLAG_LONG;
 329           else if (spec->info.is_short)
 330             spec->data_arg_type = PA_INT|PA_FLAG_SHORT;
 331           else
 332             spec->data_arg_type = PA_INT;
 333           break;
 334         case L'e':
 335         case L'E':
 336         case L'f':
 337         case L'g':
 338         case L'G':
 339           if (spec->info.is_long_double)
 340             spec->data_arg_type = PA_DOUBLE|PA_FLAG_LONG_DOUBLE;
 341           else
 342             spec->data_arg_type = PA_DOUBLE;
 343           break;
 344         case L'c':
 345           spec->data_arg_type = PA_CHAR;
 346           break;
 347         case L'C':
 348           spec->data_arg_type = PA_WCHAR;
 349           break;
 350         case L's':
 351           spec->data_arg_type = PA_STRING;
 352           break;
 353         case L'S':
 354           spec->data_arg_type = PA_WSTRING;
 355           break;
 356         case L'p':
 357           spec->data_arg_type = PA_POINTER;
 358           break;
 359         case L'n':
 360           spec->data_arg_type = PA_INT|PA_FLAG_PTR;
 361           break;
 362
 363         case L'm':
 364         default:
 365           /* An unknown spec will consume no args.  */
 366           spec->ndata_args = 0;
 367           break;
 368         }
 369
 370       if (spec->data_arg == -1 && spec->ndata_args > 0)
 371         {
 372           /* There are args consumed, but no positional spec.
 373              Use the next sequential arg position.  */
 374           spec->data_arg = posn;
 375           posn += spec->ndata_args;
 376           nargs += spec->ndata_args;
 377         }
 378     }
 379
 380   if (spec->info.spec == L'\0')
 381     /* Format ended before this spec was complete.  */
 382     spec->end_of_fmt = spec->next_fmt = format - 1;
 383   else
 384     {
 385       /* Find the next format spec.  */
 386       spec->end_of_fmt = format;
 387       spec->next_fmt = find_spec (format, ps);
 388     }
 389
 390   return nargs;
 391 }