[tz] source code question regarding localtime.c PS

Paul Eggert eggert at cs.ucla.edu
Thu Aug 8 00:20:52 UTC 2013


On 08/07/2013 10:36 AM, Alan Barrett wrote:
> the tz project assumes that time_t is encoded in seconds,
> and assumes that time_t is either an integer or floating point type.

Hah!  I just checked, and we're behind the times.
The latest version of POSIX, POSIX.1-2013,
added the requirement that time_t must be an integer type.

At some point we should remove support for floating-point time_t,
as it hasn't been tested on any real-world platform and probably
does not work.  In the meantime I pushed the following proposed doc
change to the experimental repository.

>From 8fe97ff3bc1b1edca97286c7e6bbc03da340613f Mon Sep 17 00:00:00 2001
From: Paul Eggert <eggert at cs.ucla.edu>
Date: Wed, 7 Aug 2013 17:19:22 -0700
Subject: [PATCH] * Theory: Document time_t better and update for POSIX.1-2013.

Change "`" to "'"; these days, "`" and "'" are not symmetric.
---
 Theory | 125 +++++++++++++++++++++++++++++++++++++----------------------------
 1 file changed, 71 insertions(+), 54 deletions(-)

diff --git a/Theory b/Theory
index 751b12d..1b5374a 100644
--- a/Theory
+++ b/Theory
@@ -12,16 +12,13 @@ This file is in the public domain, so clarified as of
 
 ----- Time and date functions -----
 
-These time and date functions are upwards compatible with POSIX,
+These time and date functions are upwards compatible with those of POSIX,
 an international standard for UNIX-like systems.
 As of this writing, the current edition of POSIX is:
 
-  Standard for Information technology
-  -- Portable Operating System Interface (POSIX (R))
-  -- System Interfaces
-  IEEE Std 1003.1, 2004 Edition
-  <http://www.opengroup.org/online-pubs?DOC=7999959899>
-  <http://www.opengroup.org/pubs/catalog/t041.htm>
+  The Open Group Base Specifications Issue 7
+  IEEE Std 1003.1, 2013 Edition
+  <http://pubs.opengroup.org/onlinepubs/9699919799/>
 
 POSIX has the following properties and limitations.
 
@@ -34,7 +31,7 @@ POSIX has the following properties and limitations.
 
 	The POSIX TZ string takes the following form:
 
-		stdoffset[dst[offset],date[/time],date[/time]]
+		stdoffset[dst[offset][,date[/time],date[/time]]]
 
 	where:
 
@@ -45,15 +42,17 @@ POSIX has the following properties and limitations.
 		in a quoted form like "<UTC+10>"; this allows
 		"+" and "-" in the names.
 	offset
-		is of the form `[-]hh:[mm[:ss]]' and specifies the
-		offset west of UTC.  The default DST offset is one hour
-		ahead of standard time.
+		is of the form '[+-]hh:[mm[:ss]]' and specifies the
+		offset west of UTC.  'hh' may be a single digit; 0<=hh<=24.
+		The default DST offset is one hour ahead of standard time.
 	date[/time],date[/time]
 		specifies the beginning and end of DST.  If this is absent,
 		the system supplies its own rules for DST, and these can
 		differ from year to year; typically US DST rules are used.
 	time
-		takes the form `hh:[mm[:ss]]' and defaults to 02:00.
+		takes the form 'hh:[mm[:ss]]' and defaults to 02:00.
+		This is the same format as the offset, except that a
+		leading '+' or '-' is not allowed.
 	date
 		takes one of the following forms:
 		Jn (1<=n<=365)
@@ -63,8 +62,10 @@ POSIX has the following properties and limitations.
 		Mm.n.d (0[Sunday]<=d<=6[Saturday], 1<=n<=5, 1<=m<=12)
 			for the dth day of week n of month m of the year,
 			where week 1 is the first week in which day d appears,
-			and `5' stands for the last week in which day d appears
+			and '5' stands for the last week in which day d appears
 			(which may be either the 4th or 5th week).
+			Typically, this is the only useful form;
+			the n and Jn forms are rarely used.
 
 	Here is an example POSIX TZ string, for US Pacific time using rules
 	appropriate from 1987 through 2006:
@@ -95,6 +96,20 @@ POSIX has the following properties and limitations.
 
 *	POSIX requires that systems ignore leap seconds.
 
+*	The tz code attempts attempts to support all the time_t implementations
+	allowed by POSIX.  The time_t type represents a nonnegative count of
+	seconds since 1970-01-01 00:00:00 UTC, ignoring leap seconds.
+	In practice, time_t is usually a signed 64- or 32-bit integer; 32-bit
+	signed time_t values stop working after 2038-01-19 03:14:07 UTC, so
+	new implementations these days typically use a signed 64-bit integer.
+	Unsigned 32-bit integers are used on one or two platforms,
+	and 36-bit integers are also used occasionally.
+	Although POSIX.1-2013 requires time_t to be an integer type,
+	earlier POSIX versions also allowed time_t to be a floating-point type.
+	No known platforms have a floating-point time_t and although
+	the tz code attempts to support floating-point time_t this has not
+	been tested and will probably be removed at some point.
+
 These are the extensions that have been made to the POSIX functions:
 
 *	The "TZ" environment variable is used in generating the name of a file
@@ -146,6 +161,8 @@ These are the extensions that have been made to the POSIX functions:
 	environment variable; portable applications should not, however, rely
 	on this behavior since it's not the way SVR2 systems behave.)
 
+*	Negative time_t values are supported, on systems where time_t is signed.
+
 *	These functions can account for leap seconds, thanks to Bradley White.
 
 Points of interest to folks with other systems:
@@ -155,7 +172,7 @@ Points of interest to folks with other systems:
 	On such hosts, the primary use of this package
 	is to update obsolete time zone rule tables.
 	To do this, you may need to compile the time zone compiler
-	`zic' supplied with this package instead of using the system `zic',
+	'zic' supplied with this package instead of using the system 'zic',
 	since the format of zic's input changed slightly in late 1994,
 	and many vendors still do not support the new input format.
 
@@ -251,19 +268,19 @@ one example.
 Names normally have the form AREA/LOCATION, where AREA is the name
 of a continent or ocean, and LOCATION is the name of a specific
 location within that region.  North and South America share the same
-area, `America'.  Typical names are `Africa/Cairo', `America/New_York',
-and `Pacific/Honolulu'.
+area, 'America'.  Typical names are 'Africa/Cairo', 'America/New_York',
+and 'Pacific/Honolulu'.
 
 Here are the general rules used for choosing location names,
 in decreasing order of importance:
 
 	Use only valid POSIX file name components (i.e., the parts of
-		names other than `/').  Within a file name component,
-		use only ASCII letters, `.', `-' and `_'.  Do not use
+		names other than '/').  Within a file name component,
+		use only ASCII letters, '.', '-' and '_'.  Do not use
 		digits, as that might create an ambiguity with POSIX
 		TZ strings.  A file name component must not exceed 14
-		characters or start with `-'.  E.g., prefer `Brunei'
-		to `Bandar_Seri_Begawan'.
+		characters or start with '-'.  E.g., prefer 'Brunei'
+		to 'Bandar_Seri_Begawan'.
 	Do not use names that differ only in case.  Although the reference
 	        implementation is case-sensitive, some other implementations
 		are not, and they would mishandle names differing only in case.
@@ -275,51 +292,51 @@ in decreasing order of importance:
 		Otherwise these tables would become annoyingly large.
 	If a name is ambiguous, use a less ambiguous alternative;
 		e.g. many cities are named San Jose and Georgetown, so
-		prefer `Costa_Rica' to `San_Jose' and `Guyana' to `Georgetown'.
+		prefer 'Costa_Rica' to 'San_Jose' and 'Guyana' to 'Georgetown'.
 	Keep locations compact.  Use cities or small islands, not countries
 		or regions, so that any future time zone changes do not split
-		locations into different time zones.  E.g. prefer `Paris'
-		to `France', since France has had multiple time zones.
-	Use mainstream English spelling, e.g. prefer `Rome' to `Roma', and
-		prefer `Athens' to the true name (which uses Greek letters).
+		locations into different time zones.  E.g. prefer 'Paris'
+		to 'France', since France has had multiple time zones.
+	Use mainstream English spelling, e.g. prefer 'Rome' to 'Roma', and
+		prefer 'Athens' to the true name (which uses Greek letters).
 		The POSIX file name restrictions encourage this rule.
 	Use the most populous among locations in a zone,
-		e.g. prefer `Shanghai' to `Beijing'.  Among locations with
+		e.g. prefer 'Shanghai' to 'Beijing'.  Among locations with
 		similar populations, pick the best-known location,
-		e.g. prefer `Rome' to `Milan'.
-	Use the singular form, e.g. prefer `Canary' to `Canaries'.
-	Omit common suffixes like `_Islands' and `_City', unless that
-		would lead to ambiguity.  E.g. prefer `Cayman' to
-		`Cayman_Islands' and `Guatemala' to `Guatemala_City',
-		but prefer `Mexico_City' to `Mexico' because the country
+		e.g. prefer 'Rome' to 'Milan'.
+	Use the singular form, e.g. prefer 'Canary' to 'Canaries'.
+	Omit common suffixes like '_Islands' and '_City', unless that
+		would lead to ambiguity.  E.g. prefer 'Cayman' to
+		'Cayman_Islands' and 'Guatemala' to 'Guatemala_City',
+		but prefer 'Mexico_City' to 'Mexico' because the country
 		of Mexico has several time zones.
-	Use `_' to represent a space.
-	Omit `.' from abbreviations in names, e.g. prefer `St_Helena'
-		to `St._Helena'.
+	Use '_' to represent a space.
+	Omit '.' from abbreviations in names, e.g. prefer 'St_Helena'
+		to 'St._Helena'.
 	Do not change established names if they only marginally
 		violate the above rules.  For example, don't change
-		the existing name `Rome' to `Milan' merely because
+		the existing name 'Rome' to 'Milan' merely because
 		Milan's population has grown to be somewhat greater
 		than Rome's.
-	If a name is changed, put its old spelling in the `backward' file.
+	If a name is changed, put its old spelling in the 'backward' file.
 		This means old spellings will continue to work.
 
-The file `zone.tab' lists the geographical locations used to name
+The file 'zone.tab' lists the geographical locations used to name
 time zone rule files.  It is intended to be an exhaustive list
 of names for geographic regions as described above.
 
 Older versions of this package used a different naming scheme,
 and these older names are still supported.
-See the file `backward' for most of these older names
-(e.g. `US/Eastern' instead of `America/New_York').
+See the file 'backward' for most of these older names
+(e.g. 'US/Eastern' instead of 'America/New_York').
 The other old-fashioned names still supported are
-`WET', `CET', `MET', and `EET' (see the file `europe').
+'WET', 'CET', 'MET', and 'EET' (see the file 'europe').
 
 
 ----- Time zone abbreviations -----
 
 When this package is installed, it generates time zone abbreviations
-like `EST' to be compatible with human tradition and POSIX.
+like 'EST' to be compatible with human tradition and POSIX.
 Here are the general rules used for choosing time zone abbreviations,
 in decreasing order of importance:
 
@@ -346,24 +363,24 @@ in decreasing order of importance:
 		letters.
 
 	Use abbreviations that are in common use among English-speakers,
-		e.g. `EST' for Eastern Standard Time in North America.
+		e.g. 'EST' for Eastern Standard Time in North America.
 		We assume that applications translate them to other languages
 		as part of the normal localization process; for example,
-		a French application might translate `EST' to `HNE'.
+		a French application might translate 'EST' to 'HNE'.
 
 	For zones whose times are taken from a city's longitude, use the
-		traditional xMT notation, e.g. `PMT' for Paris Mean Time.
-		The only name like this in current use is `GMT'.
+		traditional xMT notation, e.g. 'PMT' for Paris Mean Time.
+		The only name like this in current use is 'GMT'.
 
 	If there is no common English abbreviation, abbreviate the English
 		translation of the usual phrase used by native speakers.
 		If this is not available or is a phrase mentioning the country
-		(e.g. ``Cape Verde Time''), then:
+		(e.g. "Cape Verde Time"), then:
 
 		When a country is identified with a single or principal zone,
-			append `T' to the country's ISO	code, e.g. `CVT' for
-			Cape Verde Time.  For summer time append `ST';
-			for double summer time append `DST'; etc.
+			append 'T' to the country's ISO	code, e.g. 'CVT' for
+			Cape Verde Time.  For summer time append 'ST';
+			for double summer time append 'DST'; etc.
 		Otherwise, take the first three letters of an English place
 			name identifying each zone and append 'T', 'ST', etc.
 			as before; e.g. 'VLAST' for VLAdivostok Summer Time.
@@ -373,10 +390,10 @@ in decreasing order of importance:
 		in some sense, asleep.
 
 Application writers should note that these abbreviations are ambiguous
-in practice: e.g. `EST' has a different meaning in Australia than
+in practice: e.g. 'EST' has a different meaning in Australia than
 it does in the United States.  In new applications, it's often better
-to use numeric UTC offsets like `-0500' instead of time zone
-abbreviations like `EST'; this avoids the ambiguity.
+to use numeric UTC offsets like '-0500' instead of time zone
+abbreviations like 'EST'; this avoids the ambiguity.
 
 
 ----- Calendrical issues -----
@@ -401,7 +418,7 @@ and (in Paris only) 1871-05-06 through 1871-05-23.
 Russia
 
 From Chris Carrier (1996-12-02):
-On 1929-10-01 the Soviet Union instituted an ``Eternal Calendar''
+On 1929-10-01 the Soviet Union instituted an "Eternal Calendar"
 with 30-day months plus 5 holidays, with a 5-day week.
 On 1931-12-01 it changed to a 6-day week; in 1934 it reverted to the
 Gregorian calendar while retaining the 6-day week; on 1940-06-27 it
-- 
1.8.1.2





More information about the tz mailing list