path: root/lib/getdate.y

%{
/* Parse a string into an internal time stamp.
   Copyright (C) 1999, 2000, 2002, 2003, 2004 Free Software Foundation, Inc.

   This program is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2, or (at your option)
   any later version.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with this program; if not, write to the Free Software Foundation,
   Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */

/* Originally written by Steven M. Bellovin <smb@research.att.com> while
   at the University of North Carolina at Chapel Hill.  Later tweaked by
   a couple of people on Usenet.  Completely overhauled by Rich $alz
   <rsalz@bbn.com> and Jim Berets <jberets@bbn.com> in August, 1990.

   Modified by Paul Eggert <eggert@twinsun.com> in August 1999 to do
   the right thing about local DST.  Also modified by Paul Eggert
   <eggert@cs.ucla.edu> in February 2004 to support
   nanosecond-resolution time stamps, and in October 2004 to support
   TZ strings in dates.  */

/* FIXME: Check for arithmetic overflow in all cases, not just
   some of them.  */

#ifdef HAVE_CONFIG_H
# include <config.h>
#endif

#include "getdate.h"

#include <alloca.h>

/* Since the code of getdate.y is not included in the Emacs executable
   itself, there is no need to #define static in this file.  Even if
   the code were included in the Emacs executable, it probably
   wouldn't do any harm to #undef it here; this will only cause
   problems if we try to write to a static variable, which I don't
   think this code needs to do.  */
#ifdef emacs
# undef static
#endif

#include <ctype.h>
#include <limits.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

#include "setenv.h"
#include "xalloc.h"

#if STDC_HEADERS || (! defined isascii && ! HAVE_ISASCII)
# define IN_CTYPE_DOMAIN(c) 1
#else
# define IN_CTYPE_DOMAIN(c) isascii (c)
#endif

#define ISSPACE(c) (IN_CTYPE_DOMAIN (c) && isspace (c))
#define ISALPHA(c) (IN_CTYPE_DOMAIN (c) && isalpha (c))
#define ISLOWER(c) (IN_CTYPE_DOMAIN (c) && islower (c))

/* ISDIGIT differs from isdigit, as follows:
   - Its arg may be any int or unsigned int; it need not be an unsigned char.
   - It's guaranteed to evaluate its argument exactly once.
   - It's typically faster.
   POSIX says that only '0' through '9' are digits.  Prefer ISDIGIT to
   isdigit unless it's important to use the locale's definition
   of `digit' even when the host does not conform to POSIX.  */
#define ISDIGIT(c) ((unsigned int) (c) - '0' <= 9)

#if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__
# define __attribute__(x)
#endif

#ifndef ATTRIBUTE_UNUSED
# define ATTRIBUTE_UNUSED __attribute__ ((__unused__))
#endif

/* Shift A right by B bits portably, by dividing A by 2**B and
   truncating towards minus infinity.  A and B should be free of side
   effects, and B should be in the range 0 <= B <= INT_BITS - 2, where
   INT_BITS is the number of useful bits in an int.  GNU code can
   assume that INT_BITS is at least 32.

   ISO C99 says that A >> B is implementation-defined if A < 0.  Some
   implementations (e.g., UNICOS 9.0 on a Cray Y-MP EL) don't shift
   right in the usual way when A < 0, so SHR falls back on division if
   ordinary A >> B doesn't seem to be the usual signed shift.  */
#define SHR(a, b)	\
  (-1 >> 1 == -1	\
   ? (a) >> (b)		\
   : (a) / (1 << (b)) - ((a) % (1 << (b)) < 0))

#define EPOCH_YEAR 1970
#define TM_YEAR_BASE 1900

#define HOUR(x) ((x) * 60)

/* An integer value, and the number of digits in its textual
   representation.  */
typedef struct
{
  long int value;
  size_t digits;
} textint;

/* An entry in the lexical lookup table.  */
typedef struct
{
  char const *name;
  int type;
  int value;
} table;

/* Meridian: am, pm, or 24-hour style.  */
enum { MERam, MERpm, MER24 };

enum { BILLION = 1000000000, LOG10_BILLION = 9 };

/* Information passed to and from the parser.  */
typedef struct
{
  /* The input string remaining to be parsed. */
  const char *input;

  /* N, if this is the Nth Tuesday.  */
  long int day_ordinal;

  /* Day of week; Sunday is 0.  */
  int day_number;

  /* tm_isdst flag for the local zone.  */
  int local_isdst;

  /* Time zone, in minutes east of UTC.  */
  long int time_zone;

  /* Style used for time.  */
  int meridian;

  /* Gregorian year, month, day, hour, minutes, seconds, and nanoseconds.  */
  textint year;
  long int month;
  long int day;
  long int hour;
  long int minutes;
  struct timespec seconds; /* includes nanoseconds */

  /* Relative year, month, day, hour, minutes, seconds, and nanoseconds.  */
  long int rel_year;
  long int rel_month;
  long int rel_day;
  long int rel_hour;
  long int rel_minutes;
  long int rel_seconds;
  long int rel_ns;

  /* Counts of nonterminals of various flavors parsed so far.  */
  bool timespec_seen;
  size_t dates_seen;
  size_t days_seen;
  size_t local_zones_seen;
  size_t rels_seen;
  size_t times_seen;
  size_t zones_seen;

  /* Table of local time zone abbrevations, terminated by a null entry.  */
  table local_time_zone_table[3];
} parser_control;

union YYSTYPE;
static int yylex (union YYSTYPE *, parser_control *);
static int yyerror (parser_control *, char *);

%}

/* We want a reentrant parser, even if the TZ manipulation and the calls to
   localtime and gmtime are not reentrant.  */
%pure-parser
%parse-param { parser_control *pc }
%lex-param { parser_control *pc }

/* This grammar has 13 shift/reduce conflicts. */
%expect 13

%union
{
  long int intval;
  textint textintval;
  struct timespec timespec;
}

%token tAGO tDST

%token <intval> tDAY tDAY_UNIT tDAYZONE tHOUR_UNIT tLOCAL_ZONE tMERIDIAN
%token <intval> tMINUTE_UNIT tMONTH tMONTH_UNIT tORDINAL
%token <intval> tSEC_UNIT tYEAR_UNIT tZONE

%token <textintval> tSNUMBER tUNUMBER
%token <timespec> tSDECIMAL_NUMBER tUDECIMAL_NUMBER

%type <intval> o_merid
%type <timespec> seconds signed_seconds unsigned_seconds

%%

spec:
    timespec
  | items
  ;

timespec:
    '@' seconds
      {
	pc->seconds = $2;
	pc->timespec_seen = true;
      }
  ;

items:
    /* empty */
  | items item
  ;

item:
    time
      { pc->times_seen++; }
  | local_zone
      { pc->local_zones_seen++; }
  | zone
      { pc->zones_seen++; }
  | date
      { pc->dates_seen++; }
  | day
      { pc->days_seen++; }
  | rel
      { pc->rels_seen++; }
  | number
  ;

time:
    tUNUMBER tMERIDIAN
      {
	pc->hour = $1.value;
	pc->minutes = 0;
	pc->seconds.tv_sec = 0;
	pc->seconds.tv_nsec = 0;
	pc->meridian = $2;
      }
  | tUNUMBER ':' tUNUMBER o_merid
      {
	pc->hour = $1.value;
	pc->minutes = $3.value;
	pc->seconds.tv_sec = 0;
	pc->seconds.tv_nsec = 0;
	pc->meridian = $4;
      }
  | tUNUMBER ':' tUNUMBER tSNUMBER
      {
	pc->hour = $1.value;
	pc->minutes = $3.value;
	pc->seconds.tv_sec = 0;
	pc->seconds.tv_nsec = 0;
	pc->meridian = MER24;
	pc->zones_seen++;
	pc->time_zone = $4.value % 100 + ($4.value / 100) * 60;
      }
  | tUNUMBER ':' tUNUMBER ':' unsigned_seconds o_merid
      {
	pc->hour = $1.value;
	pc->minutes = $3.value;
	pc->seconds = $5;
	pc->meridian = $6;
      }
  | tUNUMBER ':' tUNUMBER ':' unsigned_seconds tSNUMBER
      {
	pc->hour = $1.value;
	pc->minutes = $3.value;
	pc->seconds = $5;
	pc->meridian = MER24;
	pc->zones_seen++;
	pc->time_zone = $6.value % 100 + ($6.value / 100) * 60;
      }
  ;

local_zone:
    tLOCAL_ZONE
      { pc->local_isdst = $1; }
  | tLOCAL_ZONE tDST
      { pc->local_isdst = $1 < 0 ? 1 : $1 + 1; }
  ;

zone:
    tZONE
      { pc->time_zone = $1; }
  | tDAYZONE
      { pc->time_zone = $1 + 60; }
  | tZONE tDST
      { pc->time_zone = $1 + 60; }
  ;

day:
    tDAY
      {
	pc->day_ordinal = 1;
	pc->day_number = $1;
      }
  | tDAY ','
      {
	pc->day_ordinal = 1;
	pc->day_number = $1;
      }
  | tORDINAL tDAY
      {
	pc->day_ordinal = $1;
	pc->day_number = $2;
      }
  | tUNUMBER tDAY
      {
	pc->day_ordinal = $1.value;
	pc->day_number = $2;
      }
  ;

date:
    tUNUMBER '/' tUNUMBER
      {
	pc->month = $1.value;
	pc->day = $3.value;
      }
  | tUNUMBER '/' tUNUMBER '/' tUNUMBER
      {
	/* Interpret as YYYY/MM/DD if the first value has 4 or more digits,
	   otherwise as MM/DD/YY.
	   The goal in recognizing YYYY/MM/DD is solely to support legacy
	   machine-generated dates like those in an RCS log listing.  If
	   you want portability, use the ISO 8601 format.  */
	if (4 <= $1.digits)
	  {
	    pc->year = $1;
	    pc->month = $3.value;
	    pc->day = $5.value;
	  }
	else
	  {
	    pc->month = $1.value;
	    pc->day = $3.value;
	    pc->year = $5;
	  }
      }
  | tUNUMBER tSNUMBER tSNUMBER
      {
	/* ISO 8601 format.  YYYY-MM-DD.  */
	pc->year = $1;
	pc->month = -$2.value;
	pc->day = -$3.value;
      }
  | tUNUMBER tMONTH tSNUMBER
      {
	/* e.g. 17-JUN-1992.  */
	pc->day = $1.value;
	pc->month = $2;
	pc->year.value = -$3.value;
	pc->year.digits = $3.digits;
      }
  | tMONTH tSNUMBER tSNUMBER
      {
	/* e.g. JUN-17-1992.  */
	pc->month = $1;
	pc->day = -$2.value;
	pc->year.value = -$3.value;
	pc->year.digits = $3.digits;
      }
  | tMONTH tUNUMBER
      {
	pc->month = $1;
	pc->day = $2.value;
      }
  | tMONTH tUNUMBER ',' tUNUMBER
      {
	pc->month = $1;
	pc->day = $2.value;
	pc->year = $4;
      }
  | tUNUMBER tMONTH
      {
	pc->day = $1.value;
	pc->month = $2;
      }
  | tUNUMBER tMONTH tUNUMBER
      {
	pc->day = $1.value;
	pc->month = $2;
	pc->year = $3;
      }
  ;

rel:
    relunit tAGO
      {
	pc->rel_ns = -pc->rel_ns;
	pc->rel_seconds = -pc->rel_seconds;
	pc->rel_minutes = -pc->rel_minutes;
	pc->rel_hour = -pc->rel_hour;
	pc->rel_day = -pc->rel_day;
	pc->rel_month = -pc->rel_month;
	pc->rel_year = -pc->rel_year;
      }
  | relunit
  ;

relunit:
    tORDINAL tYEAR_UNIT
      { pc->rel_year += $1 * $2; }
  | tUNUMBER tYEAR_UNIT
      { pc->rel_year += $1.value * $2; }
  | tSNUMBER tYEAR_UNIT
      { pc->rel_year += $1.value * $2; }
  | tYEAR_UNIT
      { pc->rel_year += $1; }
  | tORDINAL tMONTH_UNIT
      { pc->rel_month += $1 * $2; }
  | tUNUMBER tMONTH_UNIT
      { pc->rel_month += $1.value * $2; }
  | tSNUMBER tMONTH_UNIT
      { pc->rel_month += $1.value * $2; }
  | tMONTH_UNIT
      { pc->rel_month += $1; }
  | tORDINAL tDAY_UNIT
      { pc->rel_day += $1 * $2; }
  | tUNUMBER tDAY_UNIT
      { pc->rel_day += $1.value * $2; }
  | tSNUMBER tDAY_UNIT
      { pc->rel_day += $1.value * $2; }
  | tDAY_UNIT
      { pc->rel_day += $1; }
  | tORDINAL tHOUR_UNIT
      { pc->rel_hour += $1 * $2; }
  | tUNUMBER tHOUR_UNIT
      { pc->rel_hour += $1.value * $2; }
  | tSNUMBER tHOUR_UNIT
      { pc->rel_hour += $1.value * $2; }
  | tHOUR_UNIT
      { pc->rel_hour += $1; }
  | tORDINAL tMINUTE_UNIT
      { pc->rel_minutes += $1 * $2; }
  | tUNUMBER tMINUTE_UNIT
      { pc->rel_minutes += $1.value * $2; }
  | tSNUMBER tMINUTE_UNIT
      { pc->rel_minutes += $1.value * $2; }
  | tMINUTE_UNIT
      { pc->rel_minutes += $1; }
  | tORDINAL tSEC_UNIT
      { pc->rel_seconds += $1 * $2; }
  | tUNUMBER tSEC_UNIT
      { pc->rel_seconds += $1.value * $2; }
  | tSNUMBER tSEC_UNIT
      { pc->rel_seconds += $1.value * $2; }
  | tSDECIMAL_NUMBER tSEC_UNIT
      { pc->rel_seconds += $1.tv_sec * $2; pc->rel_ns += $1.tv_nsec * $2; }
  | tUDECIMAL_NUMBER tSEC_UNIT
      { pc->rel_seconds += $1.tv_sec * $2; pc->rel_ns += $1.tv_nsec * $2; }
  | tSEC_UNIT
      { pc->rel_seconds += $1; }
  ;

seconds: signed_seconds | unsigned_seconds;

signed_seconds:
    tSDECIMAL_NUMBER
  | tSNUMBER
      { $$.tv_sec = $1.value; $$.tv_nsec = 0; }
  ;

unsigned_seconds:
    tUDECIMAL_NUMBER
  | tUNUMBER
      { $$.tv_sec = $1.value; $$.tv_nsec = 0; }
  ;

number:
    tUNUMBER
      {
	if (pc->dates_seen
	    && ! pc->rels_seen && (pc->times_seen || 2 < $1.digits))
	  pc->year = $1;
	else
	  {
	    if (4 < $1.digits)
	      {
		pc->dates_seen++;
		pc->day = $1.value % 100;
		pc->month = ($1.value / 100) % 100;
		pc->year.value = $1.value / 10000;
		pc->year.digits = $1.digits - 4;
	      }
	    else
	      {
		pc->times_seen++;
		if ($1.digits <= 2)
		  {
		    pc->hour = $1.value;
		    pc->minutes = 0;
		  }
		else
		  {
		    pc->hour = $1.value / 100;
		    pc->minutes = $1.value % 100;
		  }
		pc->seconds.tv_sec = 0;
		pc->seconds.tv_nsec = 0;
		pc->meridian = MER24;
	      }
	  }
      }
  ;

o_merid:
    /* empty */
      { $$ = MER24; }
  | tMERIDIAN
      { $$ = $1; }
  ;

%%

static table const meridian_table[] =
{
  { "AM",   tMERIDIAN, MERam },
  { "A.M.", tMERIDIAN, MERam },
  { "PM",   tMERIDIAN, MERpm },
  { "P.M.", tMERIDIAN, MERpm },
  { NULL, 0, 0 }
};

static table const dst_table[] =
{
  { "DST", tDST, 0 }
};

static table const month_and_day_table[] =
{
  { "JANUARY",	tMONTH,	 1 },
  { "FEBRUARY",	tMONTH,	 2 },
  { "MARCH",	tMONTH,	 3 },
  { "APRIL",	tMONTH,	 4 },
  { "MAY",	tMONTH,	 5 },
  { "JUNE",	tMONTH,	 6 },
  { "JULY",	tMONTH,	 7 },
  { "AUGUST",	tMONTH,	 8 },
  { "SEPTEMBER",tMONTH,	 9 },
  { "SEPT",	tMONTH,	 9 },
  { "OCTOBER",	tMONTH,	10 },
  { "NOVEMBER",	tMONTH,	11 },
  { "DECEMBER",	tMONTH,	12 },
  { "SUNDAY",	tDAY,	 0 },
  { "MONDAY",	tDAY,	 1 },
  { "TUESDAY",	tDAY,	 2 },
  { "TUES",	tDAY,	 2 },
  { "WEDNESDAY",tDAY,	 3 },
  { "WEDNES",	tDAY,	 3 },
  { "THURSDAY",	tDAY,	 4 },
  { "THUR",	tDAY,	 4 },
  { "THURS",	tDAY,	 4 },
  { "FRIDAY",	tDAY,	 5 },
  { "SATURDAY",	tDAY,	 6 },
  { NULL, 0, 0 }
};

static table const time_units_table[] =
{
  { "YEAR",	tYEAR_UNIT,	 1 },
  { "MONTH",	tMONTH_UNIT,	 1 },
  { "FORTNIGHT",tDAY_UNIT,	14 },
  { "WEEK",	tDAY_UNIT,	 7 },
  { "DAY",	tDAY_UNIT,	 1 },
  { "HOUR",	tHOUR_UNIT,	 1 },
  { "MINUTE",	tMINUTE_UNIT,	 1 },
  { "MIN",	tMINUTE_UNIT,	 1 },
  { "SECOND",	tSEC_UNIT,	 1 },
  { "SEC",	tSEC_UNIT,	 1 },
  { NULL, 0, 0 }
};

/* Assorted relative-time words. */
static table const relative_time_table[] =
{
  { "TOMORROW",	tDAY_UNIT,	 1 },
  { "YESTERDAY",tDAY_UNIT,	-1 },
  { "TODAY",	tDAY_UNIT,	 0 },
  { "NOW",	tDAY_UNIT,	 0 },
  { "LAST",	tORDINAL,	-1 },
  { "THIS",	tORDINAL,	 0 },
  { "NEXT",	tORDINAL,	 1 },
  { "FIRST",	tORDINAL,	 1 },
/*{ "SECOND",	tORDINAL,	 2 }, */
  { "THIRD",	tORDINAL,	 3 },
  { "FOURTH",	tORDINAL,	 4 },
  { "FIFTH",	tORDINAL,	 5 },
  { "SIXTH",	tORDINAL,	 6 },
  { "SEVENTH",	tORDINAL,	 7 },
  { "EIGHTH",	tORDINAL,	 8 },
  { "NINTH",	tORDINAL,	 9 },
  { "TENTH",	tORDINAL,	10 },
  { "ELEVENTH",	tORDINAL,	11 },
  { "TWELFTH",	tORDINAL,	12 },
  { "AGO",	tAGO,		 1 },
  { NULL, 0, 0 }
};

/* The time zone table.  This table is necessarily incomplete, as time
   zone abbreviations are ambiguous; e.g. Australians interpret "EST"
   as Eastern time in Australia, not as US Eastern Standard Time.
   You cannot rely on getdate to handle arbitrary time zone
   abbreviations; use numeric abbreviations like `-0500' instead.  */
static table const time_zone_table[] =
{
  { "GMT",	tZONE,     HOUR ( 0) },	/* Greenwich Mean */
  { "UT",	tZONE,     HOUR ( 0) },	/* Universal (Coordinated) */
  { "UTC",	tZONE,     HOUR ( 0) },
  { "WET",	tZONE,     HOUR ( 0) },	/* Western European */
  { "WEST",	tDAYZONE,  HOUR ( 0) },	/* Western European Summer */
  { "BST",	tDAYZONE,  HOUR ( 0) },	/* British Summer */
  { "ART",	tZONE,	  -HOUR ( 3) },	/* Argentina */
  { "BRT",	tZONE,	  -HOUR ( 3) },	/* Brazil */
  { "BRST",	tDAYZONE, -HOUR ( 3) },	/* Brazil Summer */
  { "NST",	tZONE,	 -(HOUR ( 3) + 30) },	/* Newfoundland Standard */
  { "NDT",	tDAYZONE,-(HOUR ( 3) + 30) },	/* Newfoundland Daylight */
  { "AST",	tZONE,    -HOUR ( 4) },	/* Atlantic Standard */
  { "ADT",	tDAYZONE, -HOUR ( 4) },	/* Atlantic Daylight */
  { "CLT",	tZONE,    -HOUR ( 4) },	/* Chile */
  { "CLST",	tDAYZONE, -HOUR ( 4) },	/* Chile Summer */
  { "EST",	tZONE,    -HOUR ( 5) },	/* Eastern Standard */
  { "EDT",	tDAYZONE, -HOUR ( 5) },	/* Eastern Daylight */
  { "CST",	tZONE,    -HOUR ( 6) },	/* Central Standard */
  { "CDT",	tDAYZONE, -HOUR ( 6) },	/* Central Daylight */
  { "MST",	tZONE,    -HOUR ( 7) },	/* Mountain Standard */
  { "MDT",	tDAYZONE, -HOUR ( 7) },	/* Mountain Daylight */
  { "PST",	tZONE,    -HOUR ( 8) },	/* Pacific Standard */
  { "PDT",	tDAYZONE, -HOUR ( 8) },	/* Pacific Daylight */
  { "AKST",	tZONE,    -HOUR ( 9) },	/* Alaska Standard */
  { "AKDT",	tDAYZONE, -HOUR ( 9) },	/* Alaska Daylight */
  { "HST",	tZONE,    -HOUR (10) },	/* Hawaii Standard */
  { "HAST",	tZONE,	  -HOUR (10) },	/* Hawaii-Aleutian Standard */
  { "HADT",	tDAYZONE, -HOUR (10) },	/* Hawaii-Aleutian Daylight */
  { "SST",	tZONE,    -HOUR (12) },	/* Samoa Standard */
  { "WAT",	tZONE,     HOUR ( 1) },	/* West Africa */
  { "CET",	tZONE,     HOUR ( 1) },	/* Central European */
  { "CEST",	tDAYZONE,  HOUR ( 1) },	/* Central European Summer */
  { "MET",	tZONE,     HOUR ( 1) },	/* Middle European */
  { "MEZ",	tZONE,     HOUR ( 1) },	/* Middle European */
  { "MEST",	tDAYZONE,  HOUR ( 1) },	/* Middle European Summer */
  { "MESZ",	tDAYZONE,  HOUR ( 1) },	/* Middle European Summer */
  { "EET",	tZONE,     HOUR ( 2) },	/* Eastern European */
  { "EEST",	tDAYZONE,  HOUR ( 2) },	/* Eastern European Summer */
  { "CAT",	tZONE,	   HOUR ( 2) },	/* Central Africa */
  { "SAST",	tZONE,	   HOUR ( 2) },	/* South Africa Standard */
  { "EAT",	tZONE,	   HOUR ( 3) },	/* East Africa */
  { "MSK",	tZONE,	   HOUR ( 3) },	/* Moscow */
  { "MSD",	tDAYZONE,  HOUR ( 3) },	/* Moscow Daylight */
  { "IST",	tZONE,	  (HOUR ( 5) + 30) },	/* India Standard */
  { "SGT",	tZONE,     HOUR ( 8) },	/* Singapore */
  { "KST",	tZONE,     HOUR ( 9) },	/* Korea Standard */
  { "JST",	tZONE,     HOUR ( 9) },	/* Japan Standard */
  { "GST",	tZONE,     HOUR (10) },	/* Guam Standard */
  { "NZST",	tZONE,     HOUR (12) },	/* New Zealand Standard */
  { "NZDT",	tDAYZONE,  HOUR (12) },	/* New Zealand Daylight */
  { NULL, 0, 0  }
};

/* Military time zone table. */
static table const military_table[] =
{
  { "A", tZONE,	-HOUR ( 1) },
  { "B", tZONE,	-HOUR ( 2) },
  { "C", tZONE,	-HOUR ( 3) },
  { "D", tZONE,	-HOUR ( 4) },
  { "E", tZONE,	-HOUR ( 5) },
  { "F", tZONE,	-HOUR ( 6) },
  { "G", tZONE,	-HOUR ( 7) },
  { "H", tZONE,	-HOUR ( 8) },
  { "I", tZONE,	-HOUR ( 9) },
  { "K", tZONE,	-HOUR (10) },
  { "L", tZONE,	-HOUR (11) },
  { "M", tZONE,	-HOUR (12) },
  { "N", tZONE,	 HOUR ( 1) },
  { "O", tZONE,	 HOUR ( 2) },
  { "P", tZONE,	 HOUR ( 3) },
  { "Q", tZONE,	 HOUR ( 4) },
  { "R", tZONE,	 HOUR ( 5) },
  { "S", tZONE,	 HOUR ( 6) },
  { "T", tZONE,	 HOUR ( 7) },
  { "U", tZONE,	 HOUR ( 8) },
  { "V", tZONE,	 HOUR ( 9) },
  { "W", tZONE,	 HOUR (10) },
  { "X", tZONE,	 HOUR (11) },
  { "Y", tZONE,	 HOUR (12) },
  { "Z", tZONE,	 HOUR ( 0) },
  { NULL, 0, 0 }
};



static int
to_hour (long int hours, int meridian)
{
  switch (meridian)
    {
    default: /* Pacify GCC.  */
    case MER24:
      return 0 <= hours && hours < 24 ? hours : -1;
    case MERam:
      return 0 < hours && hours < 12 ? hours : hours == 12 ? 0 : -1;
    case MERpm:
      return 0 < hours && hours < 12 ? hours + 12 : hours == 12 ? 12 : -1;
    }
}

static long int
to_year (textint textyear)
{
  long int year = textyear.value;

  if (year < 0)
    year = -year;

  /* XPG4 suggests that years 00-68 map to 2000-2068, and
     years 69-99 map to 1969-1999.  */
  else if (textyear.digits == 2)
    year += year < 69 ? 2000 : 1900;

  return year;
}

static table const *
lookup_zone (parser_control const *pc, char const *name)
{
  table const *tp;

  /* Try local zone abbreviations first; they're more likely to be right.  */
  for (tp = pc->local_time_zone_table; tp->name; tp++)
    if (strcmp (name, tp->name) == 0)
      return tp;

  for (tp = time_zone_table; tp->name; tp++)
    if (strcmp (name, tp->name) == 0)
      return tp;

  return NULL;
}

#if ! HAVE_TM_GMTOFF
/* Yield the difference between *A and *B,
   measured in seconds, ignoring leap seconds.
   The body of this function is taken directly from the GNU C Library;
   see src/strftime.c.  */
static long int
tm_diff (struct tm const *a, struct tm const *b)
{
  /* Compute intervening leap days correctly even if year is negative.
     Take care to avoid int overflow in leap day calculations.  */
  int a4 = SHR (a->tm_year, 2) + SHR (TM_YEAR_BASE, 2) - ! (a->tm_year & 3);
  int b4 = SHR (b->tm_year, 2) + SHR (TM_YEAR_BASE, 2) - ! (b->tm_year & 3);
  int a100 = a4 / 25 - (a4 % 25 < 0);
  int b100 = b4 / 25 - (b4 % 25 < 0);
  int a400 = SHR (a100, 2);
  int b400 = SHR (b100, 2);
  int intervening_leap_days = (a4 - b4) - (a100 - b100) + (a400 - b400);
  long int ayear = a->tm_year;
  long int years = ayear - b->tm_year;
  long int days = (365 * years + intervening_leap_days
		   + (a->tm_yday - b->tm_yday));
  return (60 * (60 * (24 * days + (a->tm_hour - b->tm_hour))
		+ (a->tm_min - b->tm_min))
	  + (a->tm_sec - b->tm_sec));
}
#endif /* ! HAVE_TM_GMTOFF */

static table const *
lookup_word (parser_control const *pc, char *word)
{
  char *p;
  char *q;
  size_t wordlen;
  table const *tp;
  bool period_found;
  bool abbrev;

  /* Make it uppercase.  */
  for (p = word; *p; p++)
    {
      unsigned char ch = *p;
      if (ISLOWER (ch))
	*p = toupper (ch);
    }

  for (tp = meridian_table; tp->name; tp++)
    if (strcmp (word, tp->name) == 0)
      return tp;

  /* See if we have an abbreviation for a month. */
  wordlen = strlen (word);
  abbrev = wordlen == 3 || (wordlen == 4 && word[3] == '.');

  for (tp = month_and_day_table; tp->name; tp++)
    if ((abbrev ? strncmp (word, tp->name, 3) : strcmp (word, tp->name)) == 0)
      return tp;

  if ((tp = lookup_zone (pc, word)))
    return tp;

  if (strcmp (word, dst_table[0].name) == 0)
    return dst_table;

  for (tp = time_units_table; tp->name; tp++)
    if (strcmp (word, tp->name) == 0)
      return tp;

  /* Strip off any plural and try the units table again. */
  if (word[wordlen - 1] == 'S')
    {
      word[wordlen - 1] = '\0';
      for (tp = time_units_table; tp->name; tp++)
	if (strcmp (word, tp->name) == 0)
	  return tp;
      word[wordlen - 1] = 'S';	/* For "this" in relative_time_table.  */
    }

  for (tp = relative_time_table; tp->name; tp++)
    if (strcmp (word, tp->name) == 0)
      return tp;

  /* Military time zones. */
  if (wordlen == 1)
    for (tp = military_table; tp->name; tp++)
      if (word[0] == tp->name[0])
	return tp;

  /* Drop out any periods and try the time zone table again. */
  for (period_found = false, p = q = word; (*p = *q); q++)
    if (*q == '.')
      period_found = true;
    else
      p++;
  if (period_found && (tp = lookup_zone (pc, word)))
    return tp;

  return NULL;
}

static int
yylex (YYSTYPE *lvalp, parser_control *pc)
{
  unsigned char c;
  size_t count;

  for (;;)
    {
      while (c = *pc->input, ISSPACE (c))
	pc->input++;

      if (ISDIGIT (c) || c == '-' || c == '+')
	{
	  char const *p;
	  int sign;
	  unsigned long int value;
	  if (c == '-' || c == '+')
	    {
	      sign = c == '-' ? -1 : 1;
	      while (c = *++pc->input, ISSPACE (c))
		continue;
	      if (! ISDIGIT (c))
		/* skip the '-' sign */
		continue;
	    }
	  else
	    sign = 0;
	  p = pc->input;
	  for (value = 0; ; value *= 10)
	    {
	      unsigned long int value1 = value + (c - '0');
	      if (value1 < value)
		return '?';
	      value = value1;
	      c = *++p;
	      if (! ISDIGIT (c))
		break;
	      if (ULONG_MAX / 10 < value)
		return '?';
	    }
	  if ((c == '.' || c == ',') && ISDIGIT (p[1]))
	    {
	      time_t s;
	      int ns;
	      int digits;
	      unsigned long int value1;

	      /* Check for overflow when converting value to time_t.  */
	      if (sign < 0)
		{
		  s = - value;
		  if (0 < s)
		    return '?';
		  value1 = -s;
		}
	      else
		{
		  s = value;
		  if (s < 0)
		    return '?';
		  value1 = s;
		}
	      if (value != value1)
		return '?';

	      /* Accumulate fraction, to ns precision.  */
	      p++;
	      ns = *p++ - '0';
	      for (digits = 2; digits <= LOG10_BILLION; digits++)
		{
		  ns *= 10;
		  if (ISDIGIT (*p))
		    ns += *p++ - '0';
		}

	      /* Skip excess digits, truncating toward -Infinity.  */
	      if (sign < 0)
		for (; ISDIGIT (*p); p++)
		  if (*p != '0')
		    {
		      ns++;
		      break;
		    }
	      while (ISDIGIT (*p))
		p++;

	      /* Adjust to the timespec convention, which is that
		 tv_nsec is always a positive offset even if tv_sec is
		 negative.  */
	      if (sign < 0 && ns)
		{
		  s--;
		  if (! (s < 0))
		    return '?';
		  ns = BILLION - ns;
		}

	      lvalp->timespec.tv_sec = s;
	      lvalp->timespec.tv_nsec = ns;
	      pc->input = p;
	      return sign ? tSDECIMAL_NUMBER : tUDECIMAL_NUMBER;
	    }
	  else
	    {
	      if (sign < 0)
		{
		  lvalp->textintval.value = - value;
		  if (0 < lvalp->textintval.value)
		    return '?';
		}
	      else
		{
		  lvalp->textintval.value = value;
		  if (lvalp->textintval.value < 0)
		    return '?';
		}
	      lvalp->textintval.digits = p - pc->input;
	      pc->input = p;
	      return sign ? tSNUMBER : tUNUMBER;
	    }
	}

      if (ISALPHA (c))
	{
	  char buff[20];
	  char *p = buff;
	  table const *tp;

	  do
	    {
	      if (p < buff + sizeof buff - 1)
		*p++ = c;
	      c = *++pc->input;
	    }
	  while (ISALPHA (c) || c == '.');

	  *p = '\0';
	  tp = lookup_word (pc, buff);
	  if (! tp)
	    return '?';
	  lvalp->intval = tp->value;
	  return tp->type;
	}

      if (c != '(')
	return *pc->input++;
      count = 0;
      do
	{
	  c = *pc->input++;
	  if (c == '\0')
	    return c;
	  if (c == '(')
	    count++;
	  else if (c == ')')
	    count--;
	}
      while (count != 0);
    }
}

/* Do nothing if the parser reports an error.  */
static int
yyerror (parser_control *pc ATTRIBUTE_UNUSED, char *s ATTRIBUTE_UNUSED)
{
  return 0;
}

/* If *TM0 is the old and *TM1 is the new value of a struct tm after
   passing it to mktime, return true if it's OK that mktime returned T.
   It's not OK if *TM0 has out-of-range members.  */

static bool
mktime_ok (struct tm const *tm0, struct tm const *tm1, time_t t)
{
  if (t == (time_t) -1)
    {
      /* Guard against falsely reporting an error when parsing a time
	 stamp that happens to equal (time_t) -1, on a host that
	 supports such a time stamp.  */
      tm1 = localtime (&t);
      if (!tm1)
	return false;
    }

  return ! ((tm0->tm_sec ^ tm1->tm_sec)
	    | (tm0->tm_min ^ tm1->tm_min)
	    | (tm0->tm_hour ^ tm1->tm_hour)
	    | (tm0->tm_mday ^ tm1->tm_mday)
	    | (tm0->tm_mon ^ tm1->tm_mon)
	    | (tm0->tm_year ^ tm1->tm_year));
}

/* A reasonable upper bound for the size of ordinary TZ strings.
   Use heap allocation if TZ's length exceeds this.  */
enum { TZBUFSIZE = 100 };

/* Return a copy of TZ, stored in TZBUF if it fits, and heap-allocated
   otherwise.  */
static char *
get_tz (char tzbuf[TZBUFSIZE])
{
  char *tz = getenv ("TZ");
  if (tz)
    {
      size_t tzsize = strlen (tz) + 1;
      tz = (tzsize <= TZBUFSIZE
	    ? memcpy (tzbuf, tz, tzsize)
	    : xmemdup (tz, tzsize));
    }
  return tz;
}

/* Parse a date/time string, storing the resulting time value into *RESULT.
   The string itself is pointed to by P.  Return true if successful.
   P can be an incomplete or relative time specification; if so, use
   *NOW as the basis for the returned time.  */
bool
get_date (struct timespec *result, char const *p, struct timespec const *now)
{
  time_t Start;
  long int Start_ns;
  struct tm const *tmp;
  struct tm tm;
  struct tm tm0;
  parser_control pc;
  struct timespec gettime_buffer;
  unsigned char c;
  bool tz_was_altered = false;
  char *tz0 = NULL;
  char tz0buf[TZBUFSIZE];
  bool ok = true;

  if (! now)
    {
      if (gettime (&gettime_buffer) != 0)
	return false;
      now = &gettime_buffer;
    }

  Start = now->tv_sec;
  Start_ns = now->tv_nsec;

  tmp = localtime (&now->tv_sec);
  if (! tmp)
    return false;

  while (c = *p, ISSPACE (c))
    p++;

  if (strncmp (p, "TZ=\"", 4) == 0)
    {
      char const *tzbase = p + 4;
      size_t tzsize = 1;
      char const *s;

      for (s = tzbase; *s; s++, tzsize++)
	if (*s == '\\')
	  {
	    s++;
	    if (! (*s == '\\' || *s == '"'))
	      break;
	  }
	else if (*s == '"')
	  {
	    char *z;
	    char *tz1;
	    char tz1buf[TZBUFSIZE];
	    bool large_tz = TZBUFSIZE < tzsize;
	    bool setenv_ok;
	    tz0 = get_tz (tz0buf);
	    z = tz1 = large_tz ? xmalloc (tzsize) : tz1buf;
	    for (s = tzbase; *s != '"'; s++)
	      *z++ = *(s += *s == '\\');
	    *z = '\0';
	    setenv_ok = setenv ("TZ", tz1, 1) == 0;
	    if (large_tz)
	      free (tz1);
	    if (!setenv_ok)
	      goto fail;
	    tz_was_altered = true;
	    p = s + 1;
	  }
    }

  pc.input = p;
  pc.year.value = tmp->tm_year;
  pc.year.value += TM_YEAR_BASE;
  pc.year.digits = 4;
  pc.month = tmp->tm_mon + 1;
  pc.day = tmp->tm_mday;
  pc.hour = tmp->tm_hour;
  pc.minutes = tmp->tm_min;
  pc.seconds.tv_sec = tmp->tm_sec;
  pc.seconds.tv_nsec = Start_ns;
  tm.tm_isdst = tmp->tm_isdst;

  pc.meridian = MER24;
  pc.rel_ns = 0;
  pc.rel_seconds = 0;
  pc.rel_minutes = 0;
  pc.rel_hour = 0;
  pc.rel_day = 0;
  pc.rel_month = 0;
  pc.rel_year = 0;
  pc.timespec_seen = false;
  pc.dates_seen = 0;
  pc.days_seen = 0;
  pc.rels_seen = 0;
  pc.times_seen = 0;
  pc.local_zones_seen = 0;
  pc.zones_seen = 0;

#if HAVE_STRUCT_TM_TM_ZONE
  pc.local_time_zone_table[0].name = tmp->tm_zone;
  pc.local_time_zone_table[0].type = tLOCAL_ZONE;
  pc.local_time_zone_table[0].value = tmp->tm_isdst;
  pc.local_time_zone_table[1].name = NULL;

  /* Probe the names used in the next three calendar quarters, looking
     for a tm_isdst different from the one we already have.  */
  {
    int quarter;
    for (quarter = 1; quarter <= 3; quarter++)
      {
	time_t probe = Start + quarter * (90 * 24 * 60 * 60);
	struct tm const *probe_tm = localtime (&probe);
	if (probe_tm && probe_tm->tm_zone
	    && probe_tm->tm_isdst != pc.local_time_zone_table[0].value)
	  {
	      {
		pc.local_time_zone_table[1].name = probe_tm->tm_zone;
		pc.local_time_zone_table[1].type = tLOCAL_ZONE;
		pc.local_time_zone_table[1].value = probe_tm->tm_isdst;
		pc.local_time_zone_table[2].name = NULL;
	      }
	    break;
	  }
      }
  }
#else
#if HAVE_TZNAME
  {
# ifndef tzname
    extern char *tzname[];
# endif
    int i;
    for (i = 0; i < 2; i++)
      {
	pc.local_time_zone_table[i].name = tzname[i];
	pc.local_time_zone_table[i].type = tLOCAL_ZONE;
	pc.local_time_zone_table[i].value = i;
      }
    pc.local_time_zone_table[i].name = NULL;
  }
#else
  pc.local_time_zone_table[0].name = NULL;
#endif
#endif

  if (pc.local_time_zone_table[0].name && pc.local_time_zone_table[1].name
      && ! strcmp (pc.local_time_zone_table[0].name,
		   pc.local_time_zone_table[1].name))
    {
      /* This locale uses the same abbrevation for standard and
	 daylight times.  So if we see that abbreviation, we don't
	 know whether it's daylight time.  */
      pc.local_time_zone_table[0].value = -1;
      pc.local_time_zone_table[1].name = NULL;
    }

  if (yyparse (&pc) != 0)
    goto fail;

  if (pc.timespec_seen)
    *result = pc.seconds;
  else
    {
      if (1 < pc.times_seen || 1 < pc.dates_seen || 1 < pc.days_seen
	  || 1 < (pc.local_zones_seen + pc.zones_seen)
	  || (pc.local_zones_seen && 1 < pc.local_isdst))
	goto fail;

      tm.tm_year = to_year (pc.year) - TM_YEAR_BASE;
      tm.tm_mon = pc.month - 1;
      tm.tm_mday = pc.day;
      if (pc.times_seen || (pc.rels_seen && ! pc.dates_seen && ! pc.days_seen))
	{
	  tm.tm_hour = to_hour (pc.hour, pc.meridian);
	  if (tm.tm_hour < 0)
	    goto fail;
	  tm.tm_min = pc.minutes;
	  tm.tm_sec = pc.seconds.tv_sec;
	}
      else
	{
	  tm.tm_hour = tm.tm_min = tm.tm_sec = 0;
	  pc.seconds.tv_nsec = 0;
	}

      /* Let mktime deduce tm_isdst if we have an absolute time stamp.  */
      if (pc.dates_seen | pc.days_seen | pc.times_seen)
	tm.tm_isdst = -1;

      /* But if the input explicitly specifies local time with or without
	 DST, give mktime that information.  */
      if (pc.local_zones_seen)
	tm.tm_isdst = pc.local_isdst;

      tm0 = tm;

      Start = mktime (&tm);

      if (! mktime_ok (&tm0, &tm, Start))
	{
	  if (! pc.zones_seen)
	    goto fail;
	  else
	    {
	      /* Guard against falsely reporting errors near the time_t
		 boundaries when parsing times in other time zones.  For
		 example, suppose the input string "1969-12-31 23:00:00 -0100",
		 the current time zone is 8 hours ahead of UTC, and the min
		 time_t value is 1970-01-01 00:00:00 UTC.  Then the min
		 localtime value is 1970-01-01 08:00:00, and mktime will
		 therefore fail on 1969-12-31 23:00:00.  To work around the
		 problem, set the time zone to 1 hour behind UTC temporarily
		 by setting TZ="XXX1:00" and try mktime again.  */

	      long int time_zone = pc.time_zone;
	      long int abs_time_zone = time_zone < 0 ? - time_zone : time_zone;
	      long int abs_time_zone_hour = abs_time_zone / 60;
	      int abs_time_zone_min = abs_time_zone % 60;
	      char tz1buf[sizeof "XXX+0:00"
			  + sizeof pc.time_zone * CHAR_BIT / 3];
	      if (!tz_was_altered)
		tz0 = get_tz (tz0buf);
	      sprintf (tz1buf, "XXX%s%ld:%02d", "-" + (time_zone < 0),
		       abs_time_zone_hour, abs_time_zone_min);
	      if (setenv ("TZ", tz1buf, 1) != 0)
		goto fail;
	      tz_was_altered = true;
	      tm = tm0;
	      Start = mktime (&tm);
	      if (! mktime_ok (&tm0, &tm, Start))
		goto fail;
	    }
	}

      if (pc.days_seen && ! pc.dates_seen)
	{
	  tm.tm_mday += ((pc.day_number - tm.tm_wday + 7) % 7
			 + 7 * (pc.day_ordinal - (0 < pc.day_ordinal)));
	  tm.tm_isdst = -1;
	  Start = mktime (&tm);
	  if (Start == (time_t) -1)
	    goto fail;
	}

      if (pc.zones_seen)
	{
	  long int delta = pc.time_zone * 60;
	  time_t t1;
#ifdef HAVE_TM_GMTOFF
	  delta -= tm.tm_gmtoff;
#else
	  time_t t = Start;
	  struct tm const *gmt = gmtime (&t);
	  if (! gmt)
	    goto fail;
	  delta -= tm_diff (&tm, gmt);
#endif
	  t1 = Start - delta;
	  if ((Start < t1) != (delta < 0))
	    goto fail;	/* time_t overflow */
	  Start = t1;
	}

      /* Add relative date.  */
      if (pc.rel_year | pc.rel_month | pc.rel_day)
	{
	  int year = tm.tm_year + pc.rel_year;
	  int month = tm.tm_mon + pc.rel_month;
	  int day = tm.tm_mday + pc.rel_day;
	  if (((year < tm.tm_year) ^ (pc.rel_year < 0))
	      | ((month < tm.tm_mon) ^ (pc.rel_month < 0))
	      | ((day < tm.tm_mday) ^ (pc.rel_day < 0)))
	    goto fail;
	  tm.tm_year = year;
	  tm.tm_mon = month;
	  tm.tm_mday = day;
	  Start = mktime (&tm);
	  if (Start == (time_t) -1)
	    goto fail;
	}

      /* Add relative hours, minutes, and seconds.  On hosts that support
	 leap seconds, ignore the possibility of leap seconds; e.g.,
	 "+ 10 minutes" adds 600 seconds, even if one of them is a
	 leap second.  Typically this is not what the user wants, but it's
	 too hard to do it the other way, because the time zone indicator
	 must be applied before relative times, and if mktime is applied
	 again the time zone will be lost.  */
      {
	long int sum_ns = pc.seconds.tv_nsec + pc.rel_ns;
	long int normalized_ns = (sum_ns % BILLION + BILLION) % BILLION;
	time_t t0 = Start;
	long int d1 = 60 * 60 * pc.rel_hour;
	time_t t1 = t0 + d1;
	long int d2 = 60 * pc.rel_minutes;
	time_t t2 = t1 + d2;
	long int d3 = pc.rel_seconds;
	time_t t3 = t2 + d3;
	long int d4 = (sum_ns - normalized_ns) / BILLION;
	time_t t4 = t3 + d4;

	if ((d1 / (60 * 60) ^ pc.rel_hour)
	    | (d2 / 60 ^ pc.rel_minutes)
	    | ((t1 < t0) ^ (d1 < 0))
	    | ((t2 < t1) ^ (d2 < 0))
	    | ((t3 < t2) ^ (d3 < 0))
	    | ((t4 < t3) ^ (d4 < 0)))
	  goto fail;

	result->tv_sec = t4;
	result->tv_nsec = normalized_ns;
      }
    }

  goto done;

 fail:
  ok = false;
 done:
  if (tz_was_altered)
    ok &= (tz0 ? setenv ("TZ", tz0, 1) : unsetenv ("TZ")) == 0;
  if (tz0 != tz0buf)
    free (tz0);
  return ok;
}

#if TEST

int
main (int ac, char **av)
{
  char buff[BUFSIZ];

  printf ("Enter date, or blank line to exit.\n\t> ");
  fflush (stdout);

  buff[BUFSIZ - 1] = '\0';
  while (fgets (buff, BUFSIZ - 1, stdin) && buff[0])
    {
      struct timespec d;
      struct tm const *tm;
      if (! get_date (&d, buff, NULL))
	printf ("Bad format - couldn't convert.\n");
      else if (! (tm = localtime (&d.tv_sec)))
	{
	  long int sec = d.tv_sec;
	  printf ("localtime (%ld) failed\n", sec);
	}
      else
	{
	  int ns = d.tv_nsec;
	  printf ("%04ld-%02d-%02d %02d:%02d:%02d.%09d\n",
		  tm->tm_year + 1900L, tm->tm_mon + 1, tm->tm_mday,
		  tm->tm_hour, tm->tm_min, tm->tm_sec, ns);
	}
      printf ("\t> ");
      fflush (stdout);
    }
  return 0;
}
#endif /* TEST */