time/strptime.c

   1 /* strptime - Convert a string representation of time to a time value.
   2    Copyright (C) 1996 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1996.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Library General Public License as
   8    published by the Free Software Foundation; either version 2 of the
   9    License, or (at your option) any later version.
  10
  11    The GNU C Library is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14    Library General Public License for more details.
  15
  16    You should have received a copy of the GNU Library General Public
  17    License along with the GNU C Library; see the file COPYING.LIB.  If not,
  18    write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  19    Boston, MA 02111-1307, USA.  */
  20
  21 #include <ctype.h>
  22 #include <langinfo.h>
  23 #include <limits.h>
  24 #include <string.h>
  25 #include <time.h>
  26
  27 #include "../locale/localeinfo.h"
  28
  29
  30 #define match_char(ch1, ch2) if (ch1 != ch2) return NULL
  31 #define match_string(cs1, s2)                                                 \
  32   ({ size_t len = strlen (cs1);                                               \
  33      int result = strncasecmp (cs1, s2, len) == 0;                            \
  34      if (result) s2 += len;                                                   \
  35      result; })
  36 /* We intentionally do not use isdigit() for testing because this will
  37    lead to problems with the wide character version.  */
  38 #define get_number(from, to)                                                  \
  39   do {                                                                        \
  40     val = 0;                                                                  \
  41     if (*rp < '0' || *rp > '9')                                               \
  42       return NULL;                                                            \
  43     do {                                                                      \
  44       val *= 10;                                                              \
  45       val += *rp++ - '0';                                                     \
  46     } while (val * 10 <= to && *rp >= '0' && *rp <= '9');                     \
  47     if (val < from || val > to)                                               \
  48       return NULL;                                                            \
  49   } while (0)
  50 #define get_alt_number(from, to)                                              \
  51   do {                                                                        \
  52     const char *alts = _NL_CURRENT (LC_TIME, ALT_DIGITS);                     \
  53     val = 0;                                                                  \
  54     while (*alts != '\0')                                                     \
  55       {                                                                       \
  56         size_t len = strlen (alts);                                           \
  57         if (strncasecmp (alts, rp, len) == 0)                                 \
  58           break;                                                              \
  59         alts = strchr (alts, '\0') + 1;                                       \
  60         ++val;                                                                \
  61       }                                                                       \
  62     if (*alts == '\0')                                                        \
  63       return NULL;                                                            \
  64   } while (0)
  65 #define recursive(new_fmt)                                                    \
  66   do {                                                                        \
  67     if (*new_fmt == '\0')                                                     \
  68       return NULL;                                                            \
  69     rp = strptime (rp, new_fmt, tm);                                          \
  70     if (rp == NULL)                                                           \
  71       return NULL;                                                            \
  72   } while (0)
  73
  74
  75 char *
  76 strptime (const char *buf, const char *format, struct tm *tm)
  77 {
  78   const char *rp;
  79   const char *fmt;
  80   int cnt;
  81   size_t val;
  82   int have_I, is_pm;
  83
  84   rp = buf;
  85   fmt = format;
  86   have_I = is_pm = 0;
  87
  88   while (*fmt != '\0')
  89     {
  90       /* A white space in the format string matches 0 more or white
  91          space in the input string.  */
  92       if (isspace (*fmt))
  93         {
  94           while (isspace (*rp))
  95             ++rp;
  96           ++fmt;
  97           continue;
  98         }
  99
 100       /* Any character but `%' must be matched by the same character
 101          in the iput string.  */
 102       if (*fmt != '%')
 103         {
 104           match_char (*fmt++, *rp++);
 105           continue;
 106         }
 107
 108       ++fmt;
 109       switch (*fmt++)
 110         {
 111         case '%':
 112           /* Match the `%' character itself.  */
 113           match_char ('%', *rp++);
 114           break;
 115         case 'a':
 116         case 'A':
 117           /* Match day of week.  */
 118           for (cnt = 0; cnt < 7; ++cnt)
 119             {
 120               if (match_string (_NL_CURRENT (LC_TIME, ABDAY_1 + cnt), rp))
 121                 break;
 122               if (match_string (_NL_CURRENT (LC_TIME, DAY_1 + cnt), rp))
 123                 break;
 124             }
 125           if (cnt == 7)
 126             /* Does not match a weekday name.  */
 127             return NULL;
 128           tm->tm_wday = cnt;
 129           break;
 130         case 'b':
 131         case 'B':
 132         case 'h':
 133           /* Match month name.  */
 134           for (cnt = 0; cnt < 12; ++cnt)
 135             {
 136               if (match_string (_NL_CURRENT (LC_TIME, ABMON_1 + cnt), rp))
 137                 break;
 138               if (match_string (_NL_CURRENT (LC_TIME, MON_1 + cnt), rp))
 139                 break;
 140             }
 141           if (cnt == 12)
 142             /* Does not match a month name.  */
 143             return NULL;
 144           tm->tm_mon = cnt;
 145           break;
 146         case 'c':
 147           /* Match locale's date and time format.  */
 148           recursive (_NL_CURRENT (LC_TIME, D_T_FMT));
 149           break;
 150         case 'C':
 151           /* Match century number.  */
 152           get_number (0, 99);
 153           /* We don't need the number.  */
 154           break;
 155         case 'd':
 156         case 'e':
 157           /* Match day of month.  */
 158           get_number (1, 31);
 159           tm->tm_mday = val;
 160           break;
 161         case 'D':
 162           /* Match standard day format.  */
 163           recursive ("%m/%d/%y");
 164           break;
 165         case 'H':
 166           /* Match hour in 24-hour clock.  */
 167           get_number (0, 23);
 168           tm->tm_hour = val;
 169           have_I = 0;
 170           break;
 171         case 'I':
 172           /* Match hour in 12-hour clock.  */
 173           get_number (1, 12);
 174           tm->tm_hour = val - 1;
 175           have_I = 1;
 176           break;
 177         case 'j':
 178           /* Match day number of year.  */
 179           get_number (1, 366);
 180           tm->tm_yday = val - 1;
 181           break;
 182         case 'm':
 183           /* Match number of month.  */
 184           get_number (1, 12);
 185           tm->tm_mon = val - 1;
 186           break;
 187         case 'M':
 188           /* Match minute.  */
 189           get_number (0, 59);
 190           tm->tm_min = val;
 191           break;
 192         case 'n':
 193         case 't':
 194           /* Match any white space.  */
 195           while (isspace (*rp))
 196             ++rp;
 197           break;
 198         case 'p':
 199           /* Match locale's equivalent of AM/PM.  */
 200           if (match_string (_NL_CURRENT (LC_TIME, AM_STR), rp))
 201             break;
 202           if (match_string (_NL_CURRENT (LC_TIME, PM_STR), rp))
 203             {
 204               is_pm = 1;
 205               break;
 206             }
 207           return NULL;
 208         case 'r':
 209           recursive (_NL_CURRENT (LC_TIME, T_FMT_AMPM));
 210           break;
 211         case 'R':
 212           recursive ("%H:%M");
 213           break;
 214         case 's':
 215           {
 216             /* The number of seconds may be very high so we cannot use
 217                the `get_number' macro.  Instead read the number
 218                character for character and construct the result while
 219                doing this.  */
 220             time_t secs;
 221             if (*rp < '0' || *rp > '9')
 222               /* We need at least one digit.  */
 223               return NULL;
 224
 225             do
 226               {
 227                 secs *= 10;
 228                 secs += *rp++ - '0';
 229               }
 230             while (*rp >= '0' && *rp <= '9');
 231
 232             if (__localtime_r (&secs, tm) == NULL)
 233               /* Error in function.  */
 234               return NULL;
 235           }
 236           break;
 237         case 'S':
 238           get_number (0, 61);
 239           tm->tm_sec = val;
 240           break;
 241         case 'T':
 242           recursive ("%H:%M:%S");
 243           break;
 244         case 'u':
 245           get_number (1, 7);
 246           tm->tm_wday = val % 7;
 247           break;
 248         case 'g':
 249           get_number (0, 99);
 250           /* XXX This cannot determine any field in TM.  */
 251           break;
 252         case 'G':
 253           if (*rp < '0' || *rp > '9')
 254             return NULL;
 255           /* XXX Ignore the number since we would need some more
 256              information to compute a real date.  */
 257           do
 258             ++rp;
 259           while (*rp >= '0' && *rp <= '9');
 260           break;
 261         case 'U':
 262         case 'V':
 263         case 'W':
 264           get_number (0, 53);
 265           /* XXX This cannot determine any field in TM.  */
 266           break;
 267         case 'w':
 268           /* Match number of weekday.  */
 269           get_number (0, 6);
 270           tm->tm_wday = val;
 271           break;
 272         case 'x':
 273           recursive (_NL_CURRENT (LC_TIME, D_FMT));
 274           break;
 275         case 'X':
 276           recursive (_NL_CURRENT (LC_TIME, T_FMT));
 277           break;
 278         case 'y':
 279           /* Match year within century.  */
 280           get_number (0, 99);
 281           tm->tm_year = val;
 282           break;
 283         case 'Y':
 284           /* Match year including century number.  */
 285           get_number (0, INT_MAX);
 286           tm->tm_year = val - (val >= 2000 ? 2000 : 1900);
 287           break;
 288         case 'Z':
 289           /* XXX How to handle this?  */
 290           break;
 291         case 'E':
 292           switch (*fmt++)
 293             {
 294             case 'c':
 295               /* Match locale's alternate date and time format.  */
 296               recursive (_NL_CURRENT (LC_TIME, ERA_D_T_FMT));
 297               break;
 298             case 'C':
 299             case 'y':
 300             case 'Y':
 301               /* Match name of base year in locale's alternate
 302                  representation.  */
 303               /* XXX This is currently not implemented.  It should
 304                  use the value _NL_CURRENT (LC_TIME, ERA) but POSIX
 305                  leaves this implementation defined and we haven't
 306                  figured out how to do it yet.  */
 307               break;
 308             case 'x':
 309               recursive (_NL_CURRENT (LC_TIME, ERA_D_FMT));
 310               break;
 311             case 'X':
 312               recursive (_NL_CURRENT (LC_TIME, ERA_T_FMT));
 313               break;
 314             default:
 315               return NULL;
 316             }
 317           break;
 318         case 'O':
 319           switch (*fmt++)
 320             {
 321             case 'd':
 322             case 'e':
 323               /* Match day of month using alternate numeric symbols.  */
 324               get_alt_number (1, 31);
 325               tm->tm_mday = val;
 326               break;
 327             case 'H':
 328               /* Match hour in 24-hour clock using alternate numeric
 329                  symbols.  */
 330               get_alt_number (0, 23);
 331               tm->tm_hour = val;
 332               have_I = 0;
 333               break;
 334             case 'I':
 335               /* Match hour in 12-hour clock using alternate numeric
 336                  symbols.  */
 337               get_alt_number (1, 12);
 338               tm->tm_hour = val - 1;
 339               have_I = 1;
 340               break;
 341             case 'm':
 342               /* Match month using alternate numeric symbols.  */
 343               get_alt_number (1, 12);
 344               tm->tm_mon = val - 1;
 345               break;
 346             case 'M':
 347               /* Match minutes using alternate numeric symbols.  */
 348               get_alt_number (0, 59);
 349               tm->tm_min = val;
 350               break;
 351             case 'S':
 352               /* Match seconds using alternate numeric symbols.  */
 353               get_alt_number (0, 61);
 354               tm->tm_sec = val;
 355               break;
 356             case 'U':
 357             case 'V':
 358             case 'W':
 359               get_alt_number (0, 53);
 360               /* XXX This cannot determine any field in TM.  */
 361               break;
 362             case 'w':
 363               /* Match number of weekday using alternate numeric symbols.  */
 364               get_alt_number (0, 6);
 365               tm->tm_wday = val;
 366               break;
 367             case 'y':
 368               /* Match year within century using alternate numeric symbols.  */
 369               get_alt_number (0, 99);
 370               break;
 371             default:
 372               return NULL;
 373             }
 374           break;
 375         default:
 376           return NULL;
 377         }
 378     }
 379
 380   if (have_I && is_pm)
 381     tm->tm_hour += 12;
 382
 383   return (char *) rp;
 384 }