xref: /freebsd/contrib/tzcode/time2posix.3 (revision 3a9e6624ebeff00fb2330606acdcb194504ae05f)
1.\" This file is in the public domain, so clarified as of
2.\" 1996-06-05 by Arthur David Olson.
3.\"
4.\" $FreeBSD$
5.\"
6.Dd December 15, 2022
7.Dt TIME2POSIX 3
8.Os
9.Sh NAME
10.Nm time2posix ,
11.Nm posix2time
12.Nd convert seconds since the Epoch
13.Sh LIBRARY
14.Lb libc
15.Sh SYNOPSIS
16.In time.h
17.Ft time_t
18.Fn time2posix "time_t t"
19.Ft time_t
20.Fn posix2time "time_t t"
21.Sh DESCRIPTION
22.St -p1003.1-88
23requires the time_t value 536457599 to stand for 1986-12-31 23:59:59 UTC.
24This effectively implies that POSIX
25.Vt time_t
26values cannot include leap
27seconds and,
28therefore,
29that the system time must be adjusted as each leap occurs.
30.Pp
31If the time package is configured with leap-second support
32enabled,
33however,
34no such adjustment is needed and
35.Vt time_t
36values continue to increase over leap events
37(as a true
38.Dq "seconds since..."
39value).
40This means that these values will differ from those required by POSIX
41by the net number of leap seconds inserted since the Epoch.
42.Pp
43Typically this is not a problem as the type
44.Vt time_t
45is intended
46to be
47(mostly)
48opaque \(em
49.Vt time_t
50values should only be obtained-from and
51passed-to functions such as
52.Xr time 3 ,
53.Xr localtime 3 ,
54.Xr mktime 3
55and
56.Xr difftime 3 .
57However,
58.St -p1003.1-88
59gives an arithmetic
60expression for directly computing a
61.Vt time_t
62value from a given date/time,
63and the same relationship is assumed by some
64(usually older)
65applications.
66Any programs creating/dissecting
67.Vt time_t
68values
69using such a relationship will typically not handle intervals
70over leap seconds correctly.
71.Pp
72The
73.Fn time2posix
74and
75.Fn posix2time
76functions are provided to address this
77.Vt time_t
78mismatch by converting
79between local
80.Vt time_t
81values and their POSIX equivalents.
82This is done by accounting for the number of time-base changes that
83would have taken place on a POSIX system as leap seconds were inserted
84or deleted.
85These converted values can then be used in lieu of correcting the older
86applications,
87or when communicating with POSIX-compliant systems.
88.Pp
89The
90.Fn time2posix
91function
92is single-valued.
93That is,
94every local
95.Vt time_t
96corresponds to a single POSIX
97.Vt time_t .
98The
99.Fn posix2time
100function
101is less well-behaved:
102for a positive leap second hit the result is not unique,
103and for a negative leap second hit the corresponding
104POSIX
105.Vt time_t
106does not exist so an adjacent value is returned.
107Both of these are good indicators of the inferiority of the
108POSIX representation.
109.Pp
110The following table summarizes the relationship between a time
111T and its conversion to,
112and back from,
113the POSIX representation over the leap second inserted at the end of June,
1141993.
115.Bl -column "93/06/30" "23:59:59" "A+0" "X=time2posix(T)"
116.It Sy "DATE	TIME	T	X=time2posix(T)	posix2time(X)"
117.It "93/06/30	23:59:59	A+0	B+0	A+0"
118.It "93/06/30	23:59:60	A+1	B+1	A+1 or A+2"
119.It "93/07/01	00:00:00	A+2	B+1	A+1 or A+2"
120.It "93/07/01	00:00:01	A+3	B+2	A+3"
121.El
122.Pp
123A leap second deletion would look like...
124.Bl -column "??/06/30" "23:59:58" "A+0" "X=time2posix(T)"
125.It Sy "DATE	TIME	T	X=time2posix(T)	posix2time(X)"
126.It "??/06/30	23:59:58	A+0	B+0	A+0"
127.It "??/07/01	00:00:00	A+1	B+2	A+1"
128.It "??/07/01	00:00:01	A+2	B+3	A+2"
129.El
130.Pp
131.D1 No "[Note: posix2time(B+1) => A+0 or A+1]"
132.Pp
133If leap-second support is not enabled,
134local
135.Vt time_t
136and
137POSIX
138.Vt time_t
139values are equivalent,
140and both
141.Fn time2posix
142and
143.Fn posix2time
144degenerate to the identity function.
145.Sh "SEE ALSO"
146.Xr difftime 3 ,
147.Xr localtime 3 ,
148.Xr mktime 3 ,
149.Xr time 3
150