parse_date.re

PHP v6.0 For Linux 运行环境：Win9X/ WinME/ WinNT/ Win2K/ WinXP

		++*ptr;
	}
	return timelib_lookup_month(ptr);
}

static void timelib_eat_spaces(char **ptr)
{
	while (**ptr == ' ' || **ptr == '\t') {
		++*ptr;
	}
}

static const timelib_relunit* timelib_lookup_relunit(char **ptr)
{
	char *word;
	char *begin = *ptr, *end;
	const timelib_relunit *tp, *value = NULL;

	while (**ptr != '\0' && **ptr != ' ' && **ptr != '\t') {
		++*ptr;
	}
	end = *ptr;
	word = calloc(1, end - begin + 1);
	memcpy(word, begin, end - begin);

	for (tp = timelib_relunit_lookup; tp->name; tp++) {
		if (strcasecmp(word, tp->name) == 0) {
			value = tp;
			break;
		}
	}

	free(word);
	return value;
}

static void timelib_set_relative(char **ptr, timelib_sll amount, int behavior, Scanner *s)
{
	const timelib_relunit* relunit;

	relunit = timelib_lookup_relunit(ptr);
	switch (relunit->unit) {
		case TIMELIB_SECOND: s->time->relative.s += amount * relunit->multiplier; break;
		case TIMELIB_MINUTE: s->time->relative.i += amount * relunit->multiplier; break;
		case TIMELIB_HOUR:   s->time->relative.h += amount * relunit->multiplier; break;
		case TIMELIB_DAY:    s->time->relative.d += amount * relunit->multiplier; break;
		case TIMELIB_MONTH:  s->time->relative.m += amount * relunit->multiplier; break;
		case TIMELIB_YEAR:   s->time->relative.y += amount * relunit->multiplier; break;

		case TIMELIB_WEEKDAY:
			TIMELIB_HAVE_WEEKDAY_RELATIVE();
			TIMELIB_UNHAVE_TIME();
			s->time->relative.d += (amount > 0 ? amount - 1 : amount) * 7;
			s->time->relative.weekday = relunit->multiplier;
			s->time->relative.weekday_behavior = behavior;
			break;

		case TIMELIB_SPECIAL:
			TIMELIB_HAVE_SPECIAL_RELATIVE();
			TIMELIB_UNHAVE_TIME();
			s->time->special.type = relunit->multiplier;
			s->time->special.amount = amount;
	}
}

const static timelib_tz_lookup_table* zone_search(const char *word, long gmtoffset, int isdst)
{
	int first_found = 0;
	const timelib_tz_lookup_table  *tp, *first_found_elem = NULL;
	const timelib_tz_lookup_table  *fmp;

	if (strcasecmp("utc", word) == 0 || strcasecmp("gmt", word) == 0) {
		return timelib_timezone_utc;
	}
	
	for (tp = timelib_timezone_lookup; tp->name; tp++) {
		if (strcasecmp(word, tp->name) == 0) {
			if (!first_found) {
				first_found = 1;
				first_found_elem = tp;
				if (gmtoffset == -1) {
					return tp;
				}
			}
			if (tp->gmtoffset == gmtoffset) {
				return tp;
			}
		}
	}
	if (first_found) {
		return first_found_elem;
	}

	/* Still didn't find anything, let's find the zone solely based on
	 * offset/isdst then */
	for (fmp = timelib_timezone_fallbackmap; fmp->name; fmp++) {
		if ((fmp->gmtoffset * 3600) == gmtoffset && fmp->type == isdst) {
			return fmp;
		}
	}
	return NULL;
}

static long timelib_lookup_zone(char **ptr, int *dst, char **tz_abbr, int *found)
{
	char *word;
	char *begin = *ptr, *end;
	long  value = 0;
	const timelib_tz_lookup_table *tp;

	while (**ptr != '\0' && **ptr != ')') {
		++*ptr;
	}
	end = *ptr;
	word = calloc(1, end - begin + 1);
	memcpy(word, begin, end - begin);

	if ((tp = zone_search(word, -1, 0))) {
		value = -tp->gmtoffset / 60;
		*dst = tp->type;
		value += tp->type * 60;
		*found = 1;
	} else {
		*found = 0;
	}

	*tz_abbr = word;
	return value;
}

static long timelib_get_zone(char **ptr, int *dst, timelib_time *t, int *tz_not_found, const timelib_tzdb *tzdb)
{
	timelib_tzinfo *res;
	long            retval = 0;

	*tz_not_found = 0;

	while (**ptr == ' ' || **ptr == '\t' || **ptr == '(') {
		++*ptr;
	}
	if (**ptr == '+') {
		++*ptr;
		t->is_localtime = 1;
		t->zone_type = TIMELIB_ZONETYPE_OFFSET;
		*tz_not_found = 0;
		t->dst = 0;

		retval = -1 * timelib_parse_tz_cor(ptr);
	} else if (**ptr == '-') {
		++*ptr;
		t->is_localtime = 1;
		t->zone_type = TIMELIB_ZONETYPE_OFFSET;
		*tz_not_found = 0;
		t->dst = 0;

		retval = timelib_parse_tz_cor(ptr);
	} else {
		int found = 0;
		long offset;
		char *tz_abbr;

		t->is_localtime = 1;

		offset = timelib_lookup_zone(ptr, dst, &tz_abbr, &found);
		if (found) {
			t->zone_type = TIMELIB_ZONETYPE_ABBR;
		}
#if 0
		/* If we found a TimeZone identifier, use it */
		if (tz_name) {
			t->tz_info = timelib_parse_tzfile(tz_name);
			t->zone_type = TIMELIB_ZONETYPE_ID;
		}
#endif
		/* If we have a TimeZone identifier to start with, use it */
		if (strstr(tz_abbr, "/")) {
			if ((res = timelib_parse_tzfile(tz_abbr, tzdb)) != NULL) {
				t->tz_info = res;
				t->zone_type = TIMELIB_ZONETYPE_ID;
				found++;
			}
		}
		if (found && t->zone_type != TIMELIB_ZONETYPE_ID) {
			timelib_time_tz_abbr_update(t, tz_abbr);
		}
		free(tz_abbr);
		*tz_not_found = (found == 0);
		retval = offset;
	}
	while (**ptr == ')') {
		++*ptr;
	}
	return retval;
}

#define timelib_split_free(arg) {       \
	int i;                         \
	for (i = 0; i < arg.c; i++) {  \
		free(arg.v[i]);            \
	}                              \
	if (arg.v) {                   \
		free(arg.v);               \
	}                              \
}

static int scan(Scanner *s)
{
	uchar *cursor = s->cur;
	char *str, *ptr = NULL;
		
std:
	s->tok = cursor;
	s->len = 0;
/*!re2c
any = [\000-\377];

space = [ \t]+;
frac = "."[0-9]+;

ago = 'ago';

hour24 = [01]?[0-9] | "2"[0-3];
hour24lz = [01][0-9] | "2"[0-3];
hour12 = "0"?[1-9] | "1"[0-2];
minute = [0-5]?[0-9];
minutelz = [0-5][0-9];
second = minute | "60";
secondlz = minutelz | "60";
meridian = ([AaPp] "."? [Mm] "."?) [\000\t ];
tz = "("? [A-Za-z]{1,6} ")"? | [A-Z][a-z]+([_/][A-Z][a-z]+)+;
tzcorrection = [+-] hour24 ":"? minute?;

daysuf = "st" | "nd" | "rd" | "th";

month = "0"? [0-9] | "1"[0-2];
day   = ([0-2]?[0-9] | "3"[01]) daysuf?;
year  = [0-9]{1,4};
year2 = [0-9]{2};
year4 = [0-9]{4};

dayofyear = "00"[1-9] | "0"[1-9][0-9] | [1-2][0-9][0-9] | "3"[0-5][0-9] | "36"[0-6];
weekofyear = "0"[1-9] | [1-4][0-9] | "5"[0-3];

monthlz = "0" [1-9] | "1" [0-2];
daylz   = "0" [1-9] | [1-2][0-9] | "3" [01];

dayfull = 'sunday' | 'monday' | 'tuesday' | 'wednesday' | 'thursday' | 'friday' | 'saturday';
dayabbr = 'sun' | 'mon' | 'tue' | 'wed' | 'thu' | 'fri' | 'sat' | 'sun';
dayspecial = 'weekday' | 'weekdays';
daytext = dayfull | dayabbr | dayspecial;

monthfull = 'january' | 'february' | 'march' | 'april' | 'may' | 'june' | 'july' | 'august' | 'september' | 'october' | 'november' | 'december';
monthabbr = 'jan' | 'feb' | 'mar' | 'apr' | 'may' | 'jun' | 'jul' | 'aug' | 'sep' | 'sept' | 'oct' | 'nov' | 'dec';
monthroman = "I" | "II" | "III" | "IV" | "V" | "VI" | "VII" | "VIII" | "IX" | "X" | "XI" | "XII";
monthtext = monthfull | monthabbr | monthroman;

/* Time formats */
timetiny12 = hour12 space? meridian;
timeshort12 = hour12[:.]minutelz space? meridian;
timelong12 = hour12[:.]minute[:.]secondlz space? meridian;

timeshort24 = 't'? hour24[:.]minute;
timelong24 =  't'? hour24[:.]minute[:.]second;
iso8601long =  't'? hour24 [:.] minute [:.] second frac;

/* iso8601shorttz = hour24 [:] minutelz space? (tzcorrection | tz); */
iso8601normtz =  't'? hour24 [:.] minute [:.] secondlz space? (tzcorrection | tz);
/* iso8601longtz =  hour24 [:] minute [:] secondlz frac space? (tzcorrection | tz); */

gnunocolon       = 't'? hour24lz minutelz;
/* gnunocolontz     = hour24lz minutelz space? (tzcorrection | tz); */
iso8601nocolon   = 't'? hour24lz minutelz secondlz; 
/* iso8601nocolontz = hour24lz minutelz secondlz space? (tzcorrection | tz); */

/* Date formats */
americanshort    = month "/" day;
american         = month "/" day "/" year;
iso8601dateslash = year4 "/" monthlz "/" daylz "/"?;
dateslash        = year4 "/" month "/" day;
gnudateshorter   = year4 "-" month;
gnudateshort     = year "-" month "-" day;
iso8601date      = year4 "-" monthlz "-" daylz;
pointeddate      = day [.\t-] month [.-] year;
datefull         = day ([ \t.-])* monthtext ([ \t.-])* year;
datenoday        = monthtext ([ .\t-])* year4;
datenodayrev     = year4 ([ .\t-])* monthtext;
datetextual      = monthtext ([ .\t-])* day [,.stndrh\t ]* year;
datenoyear       = monthtext ([ .\t-])* day [,.stndrh\t ]*;
datenoyearrev    = day ([ .\t-])* monthtext;
datenocolon      = year4 monthlz daylz;

/* Special formats */
soap             = year4 "-" monthlz "-" daylz "T" hour24lz ":" minutelz ":" secondlz frac tzcorrection?;
xmlrpc           = year4 monthlz daylz "T" hour24 ":" minutelz ":" secondlz;
xmlrpcnocolon    = year4 monthlz daylz 't' hour24 minutelz secondlz;
wddx             = year4 "-" month "-" day "T" hour24 ":" minute ":" second;
pgydotd          = year4 "."? dayofyear;
pgtextshort      = monthabbr "-" daylz "-" year;
pgtextreverse    = year "-" monthabbr "-" daylz;
isoweekday       = year4 "W" weekofyear [0-7];
isoweek          = year4 "W" weekofyear;
exif             = year4 ":" monthlz ":" daylz " " hour24lz ":" minutelz ":" secondlz;

/* Common Log Format: 10/Oct/2000:13:55:36 -0700 */
clf              = day "/" monthabbr "/" year4 ":" hour24lz ":" minutelz ":" secondlz space tzcorrection;

/* Timestamp format: @1126396800 */
timestamp        = "@" "-"? [0-9]+;

/* To fix some ambiguities */
dateshortwithtimeshort12  = datenoyear timeshort12;
dateshortwithtimelong12   = datenoyear timelong12;
dateshortwithtimeshort  = datenoyear timeshort24;
dateshortwithtimelong   = datenoyear timelong24;
dateshortwithtimelongtz = datenoyear iso8601normtz;

/*
 * Relative regexps
 */
reltextnumber = 'first'|'next'|'second'|'third'|'fourth'|'fifth'|'sixth'|'seventh'|'eight'|'ninth'|'tenth'|'eleventh'|'twelfth'|'last'|'previous'|'this';
reltextunit = (('sec'|'second'|'min'|'minute'|'hour'|'day'|'week'|'fortnight'|'forthnight'|'month'|'year') 's'?) | daytext;

relnumber = ([+-]?[ \t]*[0-9]+);
relative = relnumber space? reltextunit;
relativetext = reltextnumber space reltextunit;

*/

/*!re2c
	/* so that vim highlights correctly */
	'yesterday'
	{
		DEBUG_OUTPUT("yesterday");
		TIMELIB_INIT;
		TIMELIB_HAVE_RELATIVE();
		TIMELIB_UNHAVE_TIME();

		s->time->relative.d = -1;
		TIMELIB_DEINIT;
		return TIMELIB_RELATIVE;
	}

	'now'
	{
		DEBUG_OUTPUT("now");
		TIMELIB_INIT;

		TIMELIB_DEINIT;
		return TIMELIB_RELATIVE;
	}

	'noon'
	{
		DEBUG_OUTPUT("noon");
		TIMELIB_INIT;
		TIMELIB_UNHAVE_TIME();
		TIMELIB_HAVE_TIME();
		s->time->h = 12;

		TIMELIB_DEINIT;
		return TIMELIB_RELATIVE;
	}

	'midnight' | 'today'
	{
		DEBUG_OUTPUT("midnight | today");
		TIMELIB_INIT;
		TIMELIB_UNHAVE_TIME();

		TIMELIB_DEINIT;
		return TIMELIB_RELATIVE;
	}

	'tomorrow'
	{
		DEBUG_OUTPUT("tomorrow");
		TIMELIB_INIT;
		TIMELIB_HAVE_RELATIVE();
		TIMELIB_UNHAVE_TIME();

		s->time->relative.d = 1;
		TIMELIB_DEINIT;
		return TIMELIB_RELATIVE;
	}

	timestamp
	{
		timelib_ull i;

		TIMELIB_INIT;
		TIMELIB_HAVE_RELATIVE();
		TIMELIB_UNHAVE_DATE();
		TIMELIB_UNHAVE_TIME();

		i = timelib_get_unsigned_nr((char **) &ptr, 24);
		s->time->y = 1970;
		s->time->m = 1;
		s->time->d = 1;
		s->time->h = s->time->i = s->time->s = 0;
		s->time->f = 0.0;
		s->time->relative.s += i;
		s->time->is_localtime = 1;
		s->time->zone_type = TIMELIB_ZONETYPE_OFFSET;
		s->time->z = 0;

		TIMELIB_DEINIT;
		return TIMELIB_RELATIVE;
	}

	timetiny12 | timeshort12 | timelong12
	{
		DEBUG_OUTPUT("timetiny12 | timeshort12 | timelong12");
		TIMELIB_INIT;
		TIMELIB_HAVE_TIME();
		s->time->h = timelib_get_nr((char **) &ptr, 2);
		if (*ptr == ':' || *ptr == '.') {
			s->time->i = timelib_get_nr((char **) &ptr, 2);
			if (*ptr == ':' || *ptr == '.') {
				s->time->s = timelib_get_nr((char **) &ptr, 2);
			}
		}
		s->time->h += timelib_meridian((char **) &ptr, s->time->h);
		TIMELIB_DEINIT;
		return TIMELIB_TIME12;
	}

	timeshort24 | timelong24 /* | iso8601short | iso8601norm */ | iso8601long /*| iso8601shorttz | iso8601normtz | iso8601longtz*/
	{
		int tz_not_found;
		DEBUG_OUTPUT("timeshort24 | timelong24 | iso8601long");
		TIMELIB_INIT;
		TIMELIB_HAVE_TIME();
		s->time->h = timelib_get_nr((char **) &ptr, 2);
		s->time->i = timelib_get_nr((char **) &ptr, 2);
		if (*ptr == ':' || *ptr == '.') {
			s->time->s = timelib_get_nr((char **) &ptr, 2);

			if (*ptr == '.') {
				s->time->f = timelib_get_frac_nr((char **) &ptr, 8);
			}
		}

		if (*ptr != '\0') {
			s->time->z = timelib_get_zone((char **) &ptr, &s->time->dst, s->time, &tz_not_found, s->tzdb);
			if (tz_not_found) {
				add_error(s, "The timezone could not be found in the database");
			}
		}
		TIMELIB_DEINIT;
		return TIMELIB_TIME24_WITH_ZONE;
	}

	gnunocolon
	{
		DEBUG_OUTPUT("gnunocolon");
		TIMELIB_INIT;
		switch (s->time->have_time) {
			case 0:
				s->time->h = timelib_get_nr((char **) &ptr, 2);
				s->time->i = timelib_get_nr((char **) &ptr, 2);
				s->time->s = 0;
				break;
			case 1:
				s->time->y = timelib_get_nr((char **) &ptr, 4);
				break;
			default:
				TIMELIB_DEINIT;
				add_error(s, "Double time specification");
				return TIMELIB_ERROR;
		}
		s->time->have_time++;
		TIMELIB_DEINIT;
		return TIMELIB_GNU_NOCOLON;
	}
/*
	gnunocolontz
	{
		DEBUG_OUTPUT("gnunocolontz");
		TIMELIB_INIT;
		switch (s->time->have_time) {
			case 0:
				s->time->h = timelib_get_nr((char **) &ptr, 2);
				s->time->i = timelib_get_nr((char **) &ptr, 2);
				s->time->s = 0;
				s->time->z = timelib_get_zone((char **) &ptr, &s->time->dst, s->time, s->tzdb);
				break;
			case 1:
				s->time->y = timelib_get_nr((char **) &ptr, 4);
				break;
			default:
				TIMELIB_DEINIT;
				return TIMELIB_ERROR;
		}
		s->time->have_time++;
		TIMELIB_DEINIT;
		return TIMELIB_GNU_NOCOLON_TZ;
	}
*/
	iso8601nocolon /*| iso8601nocolontz*/
	{
		int tz_not_found;
		DEBUG_OUTPUT("iso8601nocolon");
		TIMELIB_INIT;
		TIMELIB_HAVE_TIME();
		s->time->h = timelib_get_nr((char **) &ptr, 2);
		s->time->i = timelib_get_nr((char **) &ptr, 2);
		s->time->s = timelib_get_nr((char **) &ptr, 2);

		if (*ptr != '\0') {
			s->time->z = timelib_get_zone((char **) &ptr, &s->time->dst, s->time, &tz_not_found, s->tzdb);
			if (tz_not_found) {
				add_error(s, "The timezone could not be found in the database");
			}
		}
		TIMELIB_DEINIT;
		return TIMELIB_ISO_NOCOLON;
	}

	americanshort | american
	{
		DEBUG_OUTPUT("americanshort | american");
		TIMELIB_INIT;
		TIMELIB_HAVE_DATE();
		s->time->m = timelib_get_nr((char **) &ptr, 2);
		s->time->d = timelib_get_nr((char **) &ptr, 2);
		if (*ptr == '/') {
			s->time->y = timelib_get_nr((char **) &ptr, 4);
			TIMELIB_PROCESS_YEAR(s->time->y);
		}
		TIMELIB_DEINIT;
		return TIMELIB_AMERICAN;
	}

	iso8601date | iso8601dateslash | dateslash
	{
		DEBUG_OUTPUT("iso8601date | iso8601dateslash | dateslash");
		TIMELIB_INIT;
		TIMELIB_HAVE_DATE();
		s->time->y = timelib_get_nr((char **) &ptr, 4);
		s->time->m = timelib_get_nr((char **) &ptr, 2);
		s->time->d = timelib_get_nr((char **) &ptr, 2);
		TIMELIB_DEINIT;
		return TIMELIB_ISO_DATE;
	}

	gnudateshorter
	{
		DEBUG_OUTPUT("gnudateshorter");
		TIMELIB_INIT;
		TIMELIB_HAVE_DATE();
		s->time->y = timelib_get_nr((char **) &ptr, 4);
		s->time->m = timelib_get_nr((char **) &ptr, 2);