Identifiers used by implementation code which need to be publically exported. More...
#include <stdint.h>
Include dependency graph for common.h:
This graph shows which files directly or indirectly include this file:
Go to the source code of this file.
Typedefs | |
| typedef int32_t | ace_time::acetime_t |
| Type for the number of seconds from epoch. More... | |
Functions | |
| template<typename T > | |
| void | ace_time::swap (T &a, T &b) |
| Swap 2 parameters. | |
Variables | |
| const uint8_t | ace_time::kAbbrevSize = 7 + 1 |
| Size of the c-string buffer needed to hold a time zone abbreviation. More... | |
Detailed Description
Identifiers used by implementation code which need to be publically exported.
Definition in file common.h.
Typedef Documentation
◆ acetime_t
| typedef int32_t ace_time::acetime_t |
Variable Documentation
◆ kAbbrevSize
| const uint8_t ace_time::kAbbrevSize = 7 + 1 |
Size of the c-string buffer needed to hold a time zone abbreviation.
- The longest explicit abbreviation in the database (as of 2019 or so) seems to be 5 characters (https://www.timeanddate.com/time/zones/)
- The TZ database spec used to say that abbreviations are 3 to 6 characters but that wording is no longer in the document (https://data.iana.org/time-zones/theory.html#abbreviations).
- The zic(1) man page says "A time zone abbreviation has fewer than 3 or more than 6 characters. POSIX requires at least 3, and requires implementations to support at least 6". The first part of that wording makes no sense at all.
- The z specifier, added in TZDB 2024b, autogenerates the abbreviation using a [+/-][hh[mm[ss]]] pattern, which can be 7 characters long.
Let's increase the max length from 6 to 7 to handle the z. We also need one extra byte for the terminating NUL character.
