Theory revision 1.2
1# $NetBSD: Theory,v 1.2 1998/01/09 04:11:55 perry Exp $ 2from: @(#)Theory 7.5 3 4 5----- Outline ----- 6 7 Time and date functions 8 Names of time zone regions 9 Time zone abbreviations 10 11 12----- Time and date functions ----- 13 14These time and date functions are upwards compatible with POSIX.1, 15an international standard for Unix-like systems. 16As of this writing, the current edition of POSIX.1 is: 17 18 Information technology --Portable Operating System Interface (POSIX (R)) 19 -- Part 1: System Application Program Interface (API) [C Language] 20 ISO/IEC 9945-1:1996 21 ANSI/IEEE Std 1003.1, 1996 Edition 22 1996-07-12 23 24POSIX.1 has the following properties and limitations. 25 26* In POSIX.1, time display in a process is controlled by the 27 environment variable TZ. Unfortunately, the POSIX.1 TZ string takes 28 a form that is hard to describe and is error-prone in practice. 29 Also, POSIX.1 TZ strings can't deal with other (for example, Israeli) 30 daylight saving time rules, or situations where more than two 31 time zone abbreviations are used in an area. 32 33 The POSIX.1 TZ string takes the following form: 34 35 stdoffset[dst[offset],date[/time],date[/time]] 36 37 where: 38 39 std and dst 40 are 3 or more characters specifying the standard 41 and daylight saving time (DST) zone names. 42 offset 43 is of the form `[-]hh:[mm[:ss]]' and specifies the 44 offset west of UTC. The default DST offset is one hour 45 ahead of standard time. 46 date[/time],date[/time] 47 specifies the beginning and end of DST. If this is absent, 48 the system supplies its own rules for DST, and these can 49 differ from year to year; typically US DST rules are used. 50 time 51 takes the form `hh:[mm[:ss]]' and defaults to 02:00. 52 date 53 takes one of the following forms: 54 Jn (1<=n<=365) 55 origin-1 day number not counting February 29 56 n (0<=n<=365) 57 origin-0 day number counting February 29 if present 58 Mm.n.d (0[Sunday]<=d<=6[Saturday], 1<=n<=5, 1<=m<=12) 59 for the dth day of week n of month m of the year, 60 where week 1 is the first week in which day d appears, 61 and `5' stands for the last week in which day d appears 62 (which may be either the 4th or 5th week). 63 64* In POSIX.1, when a TZ value like "EST5EDT" is parsed, 65 typically the current US DST rules are used, 66 but this means that the US DST rules are compiled into each program 67 that does time conversion. This means that when US time conversion 68 rules change (as in the United States in 1987), all programs that 69 do time conversion must be recompiled to ensure proper results. 70 71* In POSIX.1, there's no tamper-proof way for a process to learn the 72 system's best idea of local wall clock. (This is important for 73 applications that an administrator wants used only at certain times-- 74 without regard to whether the user has fiddled the "TZ" environment 75 variable. While an administrator can "do everything in GMT" to get 76 around the problem, doing so is inconvenient and precludes handling 77 daylight saving time shifts--as might be required to limit phone 78 calls to off-peak hours.) 79 80* POSIX.1 requires that systems ignore leap seconds. 81 82These are the extensions that have been made to the POSIX.1 functions: 83 84* The "TZ" environment variable is used in generating the name of a file 85 from which time zone information is read (or is interpreted a la 86 POSIX); "TZ" is no longer constrained to be a three-letter time zone 87 name followed by a number of hours and an optional three-letter 88 daylight time zone name. The daylight saving time rules to be used 89 for a particular time zone are encoded in the time zone file; 90 the format of the file allows U.S., Australian, and other rules to be 91 encoded, and allows for situations where more than two time zone 92 abbreviations are used. 93 94 It was recognized that allowing the "TZ" environment variable to 95 take on values such as "America/New_York" might cause "old" programs 96 (that expect "TZ" to have a certain form) to operate incorrectly; 97 consideration was given to using some other environment variable 98 (for example, "TIMEZONE") to hold the string used to generate the 99 time zone information file name. In the end, however, it was decided 100 to continue using "TZ": it is widely used for time zone purposes; 101 separately maintaining both "TZ" and "TIMEZONE" seemed a nuisance; 102 and systems where "new" forms of "TZ" might cause problems can simply 103 use TZ values such as "EST5EDT" which can be used both by 104 "new" programs (a la POSIX) and "old" programs (as zone names and 105 offsets). 106 107* To handle places where more than two time zone abbreviations are used, 108 the functions "localtime" and "gmtime" set tzname[tmp->tm_isdst] 109 (where "tmp" is the value the function returns) to the time zone 110 abbreviation to be used. This differs from POSIX.1, where the elements 111 of tzname are only changed as a result of calls to tzset. 112 113* Since the "TZ" environment variable can now be used to control time 114 conversion, the "daylight" and "timezone" variables are no longer 115 needed. (These variables are defined and set by "tzset"; however, their 116 values will not be used by "localtime.") 117 118* The "localtime" function has been set up to deliver correct results 119 for near-minimum or near-maximum time_t values. (A comment in the 120 source code tells how to get compatibly wrong results). 121 122* A function "tzsetwall" has been added to arrange for the system's 123 best approximation to local wall clock time to be delivered by 124 subsequent calls to "localtime." Source code for portable 125 applications that "must" run on local wall clock time should call 126 "tzsetwall();" if such code is moved to "old" systems that don't 127 provide tzsetwall, you won't be able to generate an executable program. 128 (These time zone functions also arrange for local wall clock time to be 129 used if tzset is called--directly or indirectly--and there's no "TZ" 130 environment variable; portable applications should not, however, rely 131 on this behavior since it's not the way SVR2 systems behave.) 132 133* These functions can account for leap seconds, thanks to Bradley White 134 (bww@k.cs.cmu.edu). 135 136Points of interest to folks with other systems: 137 138* This package is already part of many POSIX-compliant hosts, 139 including BSD, HP, Linux, Network Appliance, SCO, SGI, and Sun. 140 On such hosts, the primary use of this package 141 is to update obsolete time zone rule tables. 142 To do this, you may need to compile the time zone compiler 143 `zic' supplied with this package instead of using the system `zic', 144 since the format of zic's input changed slightly in late 1994, 145 and many vendors still do not support the new input format. 146 147* The Unix Version 7 "timezone" function is not present in this package; 148 it's impossible to reliably map timezone's arguments (a "minutes west 149 of GMT" value and a "daylight saving time in effect" flag) to a 150 time zone abbreviation, and we refuse to guess. 151 Programs that in the past used the timezone function may now examine 152 tzname[localtime(&clock)->tm_isdst] to learn the correct time 153 zone abbreviation to use. Alternatively, use 154 localtime(&clock)->tm_zone if this has been enabled. 155 156* The 4.2BSD gettimeofday function is not used in this package. 157 This formerly let users obtain the current UTC offset and DST flag, 158 but this functionality was removed in later versions of BSD. 159 160* In SVR2, time conversion fails for near-minimum or near-maximum 161 time_t values when doing conversions for places that don't use GMT. 162 This package takes care to do these conversions correctly. 163 164The functions that are conditionally compiled if STD_INSPIRED is defined 165should, at this point, be looked on primarily as food for thought. They are 166not in any sense "standard compatible"--some are not, in fact, specified in 167*any* standard. They do, however, represent responses of various authors to 168standardization proposals. 169 170Other time conversion proposals, in particular the one developed by folks at 171Hewlett Packard, offer a wider selection of functions that provide capabilities 172beyond those provided here. The absence of such functions from this package 173is not meant to discourage the development, standardization, or use of such 174functions. Rather, their absence reflects the decision to make this package 175contain valid extensions to POSIX.1, to ensure its broad 176acceptability. If more powerful time conversion functions can be standardized, 177so much the better. 178 179 180----- Names of time zone rule files ----- 181 182The names of this package's installed time zone rule files are chosen to 183help minimize possible future incompatibilities due to political events. 184Ordinarily, names of countries are not used, to avoid incompatibilities 185when countries change their name (e.g. Zaire->Congo) or 186when locations change countries (e.g. Hong Kong from UK colony to China). 187 188Names normally have the form AREA/LOCATION, where AREA is the name 189of a continent or ocean, and LOCATION is the name of a specific 190location within that region. North and South America share the same 191area, `America'. Typical names are `Africa/Cairo', `America/New_York', 192and `Pacific/Honolulu'. 193 194Here are the general rules used for choosing location names, 195in decreasing order of importance: 196 197 Use only valid Posix file names. Use only Ascii letters, digits, `.', 198 `-' and `_'. Do not exceed 14 characters or start with `-'. 199 E.g. prefer `Brunei' to `Bandar_Seri_Begawan'. 200 Include at least one location per time zone rule set per country. 201 One such location is enough. 202 If all the clocks in a country's region have agreed since 1970, 203 don't bother to include more than one location 204 even if subregions' clocks disagreed before 1970. 205 Otherwise these tables would become annoyingly large. 206 If a name is ambiguous, use a less ambiguous alternative; 207 e.g. many cities are named San Jose and Georgetown, so 208 prefer `Costa_Rica' to `San_Jose' and `Guyana' to `Georgetown'. 209 Keep locations compact. Use cities or small islands, not countries 210 or regions, so that any future time zone changes do not split 211 locations into different time zones. E.g. prefer `Paris' 212 to `France', since France has had multiple time zones. 213 Use traditional English spelling, e.g. prefer `Rome' to `Roma', and 214 prefer `Athens' to the true name (which uses Greek letters). 215 The Posix file name restrictions encourage this rule. 216 Use the most populous among locations in a country's time zone, 217 e.g. prefer `Shanghai' to `Beijing'. Among locations with 218 similar populations, pick the best-known location, 219 e.g. prefer `Rome' to `Milan'. 220 Use the singular form, e.g. prefer `Canary' to `Canaries'. 221 Omit common suffixes like `_Islands' and `_City', unless that 222 would lead to ambiguity. E.g. prefer `Cayman' to 223 `Cayman_Islands' and `Guatemala' to `Guatemala_City', 224 but prefer `Mexico_City' to `Mexico' because the country 225 of Mexico has several time zones. 226 Use `_' to represent a space. 227 Omit `.' from abbreviations in names, e.g. prefer `St_Helena' 228 to `St._Helena'. 229 230The file `zone.tab' lists the geographical locations used to name 231time zone rule files. 232 233Older versions of this package used a different naming scheme, 234and these older names are still supported. 235See the file `backwards' for most of these older names 236(e.g. `US/Eastern' instead of `America/New_York'). 237The other old-fashioned names still supported are 238`WET', `CET', `MET', `EET' (see the file `europe'), 239and `Factory' (see the file `factory'). 240 241 242----- Time zone abbreviations ----- 243 244When this package is installed, it generates time zone abbreviations 245like `EST' to be compatible with human tradition and POSIX.1. 246Here are the general rules used for choosing time zone abbreviations, 247in decreasing order of importance: 248 249 Use abbreviations that consist of 3 or more upper-case Ascii letters, 250 except use "___" for locations while uninhabited. 251 Posix.1 requires at least 3 characters, and the restriction to 252 upper-case Ascii letters follows most traditions. 253 Previous editions of this database also used characters like 254 ' ' and '?', but these characters have a special meaning to 255 the shell and cause commands like 256 set `date` 257 to have unexpected effects. In theory, the character set could 258 be !%./@A-Z^_a-z{}, but these tables use only upper-case 259 Ascii letters (and "___"). 260 Use abbreviations that are in common use among English-speakers, 261 e.g. `EST' for Eastern Standard Time in North America. 262 We assume that applications translate them to other languages 263 as part of the normal localization process; for example, 264 a French application might translate `EST' to `HNE'. 265 For zones whose times are taken from a city's longitude, use the 266 traditional xMT notation, e.g. `PMT' for Paris Mean Time. 267 The only name like this in current use is `GMT'. 268 If there is no common English abbreviation, abbreviate the English 269 translation of the usual phrase used by native speakers. 270 If this is not available or is a phrase mentioning the country 271 (e.g. ``Cape Verde Time''), then: 272 273 When a country has a single or principal time zone region, 274 append `T' to the country's ISO code, e.g. `CVT' for 275 Cape Verde Time. For summer time append `ST'; 276 for double summer time append `DST'; etc. 277 When a country has multiple time zones, take the first three 278 letters of an English place name identifying each zone 279 and then append `T', `ST', etc. as before; 280 e.g. `VLAST' for VLAdivostok Summer Time. 281 282Application writers should note that these abbreviations are ambiguous 283in practice: e.g. `EST' has a different meaning in Australia than 284it does in the United States. In new applications, it's often better 285to use numeric GMT offsets like `-0500' instead of time zone 286abbreviations like `EST'; this avoids the ambiguity. 287