A quick test area for code to be potentially integrated into core's Datish.
Currently supports all formatters specified by POSIX.
| Code | Function | Width | Padding | Source |
|---|---|---|---|---|
%a |
name of day (abbreviated) | 0 | none | POSIX✧ |
%A |
name of day | 0 | none | POSIX✧ |
%b |
name of month (abbreviated) | 0 | none | POSIX✧ |
%B |
name of month | 0 | none | POSIX✧ |
%c |
short for %a %b %e %H:%M:%S %Y |
0 | none | POSIX✧ |
%C |
century | 2 | zeros | POSIX |
%d |
short for %m/%d/%y |
0 | zeros | POSIX |
%e |
day of month | 2 | spaces | POSIX |
%f |
microseconds | 6 | zeros | Python |
%F |
short for %+4Y-%m-%d |
0 | none | POSIX |
%g |
last two digits of week-year | 2 | zeros | POSIX |
%G |
week-year | 4 | zeros | POSIX |
%h |
abbr. name of month, alias of %b |
0 | none | POSIX✧ |
%H |
Hour (0–23 based) | 2 | zeros | POSIX |
%I |
Hour (1–12 based) | 2 | zeros | POSIX |
%j |
day of year (1–366) | 3 | zeros | POSIX |
%k |
Hour (0–23 based) | 2 | spaces | GNU/Ruby/Perl |
%l |
Hour (1–12 based) | 2 | spaces | GNU/Ruby/Perl |
%L |
milliseconds | 6 | zeros | Perl |
%m |
month (1–12) | 2 | zeros | POSIX |
%M |
minute (0–60) | 2 | zeros | POSIX |
%n |
literal newline | 0 | none | POSIX |
%N |
fractional seconds (precision=width) | 9 | none | Perl |
%p |
AM/PM | 2 | none | POSIX |
%P |
am/pm | 2 | none | Ruby |
%r |
short for %I:%M:%S %p |
0 | none | POSIX✧ |
%R |
short for %H:%M |
0 | none | GNU |
%s |
seconds since UNIX epoch | 0 | none | GNU |
%S |
seconds (0–61) | 2 | zeros | POSIX |
%t |
literal tab | 0 | none | POSIX |
%T |
short for %H:%M:%S |
0 | none | POSIX |
%u |
weekday (1–7, Monday = 1) | 0 | none | POSIX |
%U |
week (0–53, Sunday-based, 1st Sun. = 1) | 0 | none | POSIX |
%V |
week (1–53, Monday-based, 1st Thu. = 1) | 0 | none | POSIX |
%w |
weekday (0–6, Sunday = 0) | 0 | none | POSIX |
%W |
week (0–53, Monday-based, 1st Mon. = 1) | 0 | none | POSIX |
%x |
short for %m/%d/%y |
0 | none | POSIX✧ |
%X |
short for %H:%M:%S |
0 | none | POSIX✧ |
%Y |
year | 4 | zero | POSIX |
%y |
year (last two digits) | 2 | zero | POSIX |
%z |
UTC-style timezone offset | 4 | zero | POSIX |
%Z |
Timezone (or blank if unavailable) | 4 | zero | POSIX |
%+ |
short for %a %b %e %H:%M:%S %Z %Y |
0 | zero | POSIX |
%% |
literal percent | 0 | zero | POSIX |
Sources indicated with a ✧ are locale-based ones, and are fixed as they exclusively use the POSIX locale.
The style %+ is seen in some implementations (Android, Apple, GNU), which conflicts with the POSIX modifier +. It can only be obtained by way of some other modifier, like %-+.
| Flag | Effect |
|---|---|
0 |
Pad to minimum width with zeros |
_ |
Pad to minimum width with spaces |
+ |
Add a + if longer than minimum width, and −(U+2212) if negative |
- |
Do not pad (overrides minimum width to 0) |
^ |
Uppercase (overrides minimum width to 0) |
# |
Lowercase (overrides minimum width to 0) |
Only 0 and + are specified in POSIX, but the others are quite common in different implementations. # is generally a context-sensitive one, but we generalize it to lowercase.
| Flag | Effect |
|---|---|
E |
Use locale-specified variations (ignored) |
O |
Use locale-specified, numbering system (ignored) |
Modifiers are ignored because we assume the POSIX locale (effectively equivalent to en-US).
Per the POSIX standard for strftime (upon which this is based)
The results are unspecified if more than one flag character is specified, a flag character is specified without a minimum field width; a minimum field width is specified without a flag character; a modifier is specified with a flag or with a minimum field width; or if a minimum field width is specified for any conversion specifier other than C, F, G, or Y
Many implementations have come to a consensus (for ease of implementation) that the flags, minimum widths, and modifiers are all independently set, and each conversion specifier has a sort of "default" flag/width. This is the approach used and default values are indicated in the table.
Because %Z requires access to the timezone as a string (which neither DateTime nor $*TZ provide), Z uses a crude heuristic. It polls .timezone and checks to see if there is a letter in its stringified form. A module like DateTime::Timezones overrides that to give an IntStr where the integral form is the offset, and the string is the Olson ID. Per the standard, if the timezone name (per the /<alpha>/ test) is unavailable, it results in blank output.
It is possible in the future that an era formatter may be included (in lieu of %EY), and we may wish anticipate (even if in a design-by-contracty way) potential attributes like .era, season, or .cycle to encourage module authors to use them in case POSIX adopts a formatter for them.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent