Here is complete documention regarding the directives that strtime supports.
Directives can be modified using special characters.
An unmodified directive looks like %d. A modified directive looks like %-d or %_d or %:d or %^B.
-Don't pad; i.e. produce1when normally01would be produced._Pad with spaces; i.e. produce1instead of01.^Switch case (make lowercase if fully upper, uppercase otherwise), or unsigned year %Y for use with era.:Timezone with ':' delimiter e.g. "+00:00", or number with ordinal e.g. "1st".
The only cases in which a directive with any or no modifier cannot parse a timestamp produced with the output of any other modifier (i.e. "%d %m %Y" successfully parses the output of "-d %_m %Y") are:
:is used to produce an ordinal number; e.g.%:dcan parse the output of%:dbut not the output of%d,%-d, or%_d. Neither can these differently-modified directives parse the output of%:d.^is used to produce an unsigned year. Either%Yor%^Ycan correctly parse the output of%^Y, but only when an era directive%#is also present in the timestamp.
%aabbreviated weekday name, e.g. "Mon", "Tue"%Afull weekday name, e.g. "Monday", "Tuesday"%babbreviated month name, e.g. "Jan", "Feb"%Bfull month name, e.g. "January", "February"%cdate and time; same as%a %b %e %H:%M:%S %Y%Ccentury, i.e. "20" when the year is "2018"%dday of the month (01-31)%Ddate; same as%m/%d/%y%espace-padded numeric day of the month; same as%_d%fmicrosecond in second (000000-999999)%FISO 8601 date format; same as%Y-%m-%d%gtwo-digit ISO 8601 week year, i.e. "18" when the ISO week year is "2018" (00-99)%Gfull ISO 8601 week year%habbreviated month name; same as%b%Hhour number (00-23)%Ihour number (01-12)%jday of the year (000-366)%khour; same as%H(00-23)%lhour; same as%I(01-12)%Lmillisecond in second (000-999)%mmonth number (00-12)%Mminute (00-59)%nnewline character "\n"%puppercase AM/PM%Plowercase am/pm%Qnumber of microseconds since epoch%r12-hour time; same as%I:%M:%S %p%R24-hour time; same as%H:%M%snumber of seconds since epoch%Ssecond (normally 00-59; strictly 00-61)%ttab character "\t"%T24-hour time; same as%H:%M:%S%vVMS date; same as%e-%b-%Y%VISO 8601 week number (01-53)%uweekday number, starting with Monday (1-7)%Uweek number, counting from the first Sunday (0-53)%wweekday number, starting with Sunday (0-6)%Wweek number, counting from the first Monday (0-53)%xdate; same as%Dand%m/%d/%y%X24-hour time; same as%Tand%H:%M:%S%ytwo-digit year; i.e. "18" when the year is "2018" (00-99)%Yfull year number%ztimezone offset, e.g. "+0100"%Ztimezone name or offset, e.g. "UTC" or "+0100"%+date; same as%a %b %e %H:%M:%S %Z %Y%#era name, i.e. "CE" or "BCE"%%literal "%" character
Abbreviated weekday name. Using the default English settings these names are "Sun", "Mon", "Tue", "Wed", "Thu", "Fri", and "Sat".
The weekday names can be changed by passing an options object to strftime or strptime which includes a shortWeekdayNames attribute. The value of this attribute must be an array containing seven abbreviated weekday names, beginning with Sunday and ending with Saturday.
The modified directive %^a produces switched-case outputs. With the default English settings, this means the output will be "SUN", "MON", "TUE", and so on.
The %a directive, when used in a format for strptime, is also able to parse the outputs of the full weekday name directive %A. Parsing is case-insensitive. This means that, for example, the string "mOnDay" will be correctly parsed no matter whether the parsing directive is modified or not.
Full weekday name. Using the default English settings these names are "Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", and "Saturday".
The weekday names can be changed by passing an options object to strftime or strptime which includes a longWeekdayNames attribute. The value of this attribute must be an array containing seven full weekday names, beginning with Sunday and ending with Saturday.
The modified directive %^A produces switched-case outputs. With the default English settings, this means the output will be "SUNDAY", "MONDAY", "TUESDAY", and so on.
The %A directive, when used in a format for strptime, is also able to parse the outputs of the abbreviated weekday name directive %a. Parsing is case-insensitive. This means that, for example, the string "mOnDay" will be correctly parsed no matter whether the parsing directive is modified or not.
Abbreviated month name. Using the default English settings these names are "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", and "Dec".
The month names can be changed by passing an options object to strftime or strptime which includes a shortMonthNames attribute. The value of this attribute must be an array containing twelve abbreviated month names, beginning with January and ending with December.
The modified directive %^b produces switched-case outputs. With the default English settings, this means the output will be "JAN", "FEB", "MAR", and so on.
The %b directive, when used in a format for strptime, is also able to parse the outputs of the full month name directive %B. Parsing is case-insensitive. This means that, for example, the string "jANuARy" will be correctly parsed no matter whether the parsing directive is modified or not.
Full month name. Using the default English settings these names are "January", "February", "March", "April", "May", "June", "July", "August", "September", "October", "November", and "December".
The month names can be changed by passing an options object to strftime or strptime which includes a longMonthNames attribute. The value of this attribute must be an array containing twelve full month names, beginning with January and ending with December.
The modified directive %^B produces switched-case outputs. With the default English settings, this means the output will be "JANUARY", "FEBRUARY", "MARCH", and so on.
The %B directive, when used in a format for strptime, is also able to parse the outputs of the abbreviated month name directive %b. Parsing is case-insensitive. This means that, for example, the string "jANuARy" will be correctly parsed no matter whether the parsing directive is modified or not.
This directive is rewritten as %a %b %e %H:%M:%S %Y.
Century number, e.g. 20 when the year is 2018. The number is padded with zeroes if it is fewer than two digits long. In most practical use cases, this number will be either 19 or 20. However, it is possible for the century number to be outside that range, or even for it to be negative.
Typically, this directive is used in combination with %y, the two-digit year directive. It is not safe to use this directive in combination with the ISO week year directive %g; doing so will cause the information in the %C century directive to be ignored.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
Day of the month. The number is padded with zeroes if it is fewer than two digits long. The day number is always in the range 01 through 31.
The day of the month can be written using either the %d or %e directives. Either directive is able to parse day number output written using the other directive.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
This directive is rewritten as %m/%d/%y.
Day of the month. The number is padded with spaces if it is fewer than two digits long. The day number is always in the range 1 through 31.
The day of the month can be written using either the %d or %e directives. Either directive is able to parse day number output written using the other directive.
This numeric directive can be modified using the no-padding - and ordinal : modifiers. To get the day number padded with zeroes instead of with spaces, use the %d directive instead.
Microsecond in second. The number is padded with zeroes if it is fewer than six digits long. The microsecond number is always in the range 000000 through 999999.
Realistically, since at the time of writing JavaScript Date objects do not offer any better than millisecond precision, writing a timestamp with a microsecond directive will give only three significant digits followed by zeroes (e.g. 123000 would be written) and parsing a timestamp with a microsecond %f directive will erase the trailing three significant digits.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
This directive is rewritten as %Y-%m-%d.
Two-digit ISO week year, e.g. 18 when the ISO week year is 2018. The number is padded with zeroes if it is fewer than two digits long. Normally, this directive should be used in combination with the ISO week number directive %V and the weekday number directive %u. Please note that the ISO week year may differ from the calendar year %y, %Y for days of the year before January 4th.
When parsing a two-digit ISO week year, numbers less than or equal to 68 are considered to belong to the 21st century (e.g. 2068) whereas numbers greater than 68 are considered to belong to the 20th century (e.g. 1969). To avoid the possibility of reading the wrong year number from a timestamp outside the range 1969-2068, it is better to always write timestamps using the full ISO year directive %G instead of the two-digit year %g.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
You can refer to the ISO week date article on Wikipedia for more information about ISO 8601 week dates.
Full ISO week year. The number is padded with zeroes if it is fewer than four digits long. Normally, this directive should be used in combination with the ISO week number directive %V and the weekday number directive %u. Please note that the ISO week year may differ from the calendar year %y, %Y for days of the year before January 4th. Note also that it is possible for the ISO week year to be negative.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
You can refer to the ISO week date article on Wikipedia for more information about ISO 8601 week dates.
This directive is identical to the abbreviated month name %b directive.
Hour number on a 24-hour clock. The number is padded with zeroes if it is fewer than two digits long. The hour number is always in the range 00 through 23.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
Hour number on a 12-hour clock. The number is padded with zeroes if it is fewer than two digits long. The hour number is always in the range 01 through 12. Normally, this directive should be used in combination with the AM/PM directive %p or %P.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
Ordinal day number in the year. The number is padded with zeroes if it is fewer than three digits long. The day number is always in the range 000 through 366.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
This directive is identical to the hour number (24-hour) %H directive.
This directive is identical to the hour number (12-hour) %I directive.
Millisecond in second. The number is padded with zeroes if it is fewer than three digits long. The microsecond number is always in the range 000 through 999.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
Month number. The number is padded with zeroes if it is fewer than two digits long. The month number is always in the range 01 through 12.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
Minute number. The number is padded with zeroes if it is fewer than two digits long. The minute number is always in the range 00 through 59.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
This directive literally writes the newline character \n when formatting a date with strftime and matches the same newline character when parsing a timestamp using strptime.
Uppercase AM/PM text. Using the default English settings these strings are "AM" for times before noon and "PM" for noon and times after noon but before midnight. Normally, this directive is used in combination with the 12-hour hour number directive %I.
The AM/PM strings can be changed by passing an options object to strftime or strptime which includes a meridiemNames attribute. The value of this attribute must be an array containing two elements, an AM equivalent and then a PM equivalent. The strings given here are written with their original capitalization when using this %p directive.
The modified directive %^p produces switched-case outputs. With the default English settings, this means the output will be "am" or "pm". Note that, with the default English settings, this is the same result as using the lowercase am/pm %P directive.
The %p directive, when used in a format for strptime, is also able to parse the outputs of the lowercase am/pm directive %P. Parsing is case-insensitive. This means that, for example, the string "Pm" will be correctly parsed no matter whether the parsing directive is modified or not.
Lowercase am/pm text. Using the default English settings these strings are "am" for times before noon and "pm" for noon and times after noon but before midnight. Normally, this directive is used in combination with the 12-hour hour number directive %I.
The am/pm strings can be changed by passing an options object to strftime or strptime which includes a meridiemNames attribute. The value of this attribute must be an array containing two elements, an AM equivalent and then a PM equivalent. The strings given here are written after conversion to lowercase when using this %P directive.
The modified directive %^P produces switched-case outputs. With the default English settings, this means the output will be "AM" or "PM". Note that, with the default English settings, this is the same result as using the uppercase AM/PM %p directive.
The %P directive, when used in a format for strptime, is also able to parse the outputs of the uppercase AM/PM directive %o. Parsing is case-insensitive. This means that, for example, the string "Pm" will be correctly parsed no matter whether the parsing directive is modified or not.
The number of microseconds that have elapsed since 00:00:00 on 1 January 1970 in the timezone that the timestamp applies to.
Realistically, since at the time of writing JavaScript Date objects do not offer any better than millisecond precision, writing a timestamp with a microsecond directive will be followed by three zeroes, and parsing a timestamp with a microseconds %Q directive will erase the trailing three significant digits.
This numeric directive can be modified using the ordinal : modifier. However, since it is not padded, the no-padding - and space-padding _ modifiers are not applicable to this directive.
This directive is rewritten as %I:%M:%S %p.
This directive is rewritten as %H:%M.
The number of seconds that have elapsed since 00:00:00 on 1 January 1970 in the timezone that the timestamp applies to.
This numeric directive can be modified using the ordinal : modifier. However, since it is not padded, the no-padding - and space-padding _ modifiers are not applicable to this directive.
Second number. The number is padded with zeroes if it is fewer than two digits long. The second number is normally in the range 00 through 59 and always in the range 00 through 61.
The second number may be 60 in the case of a leap second, which occur every few years. The posix standard originally allowed for the possibility of double leap seconds, and this is why a value of 61 is also accepted. However, there have never been two successive leap seconds in the past and this is unlikely to occur in the future.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
This directive literally writes the horizontal tab character \t when formatting a date with strftime and matches the same tab character when parsing a timestamp using strptime.
This directive is rewritten as %H:%M:%S.
This directive is rewritten as %e-%b-%Y.
ISO 8601 week number. Normally, this directive should be used in combination with the ISO week year directive %G and the weekday number directive %u. This number is always in the range 01 through 53. It is padded with zeroes if it is fewer than two digits long.
You can refer to the ISO week date article on Wikipedia for more information about ISO 8601 week dates.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
The weekday number, starting with Monday and ending with Sunday. The number is always in the range 1 through 7.
Note that you will get the same results parsing a weekday number with the %u directive regardless of whether the timestamp was written with the %u or the %w weekday number directive.
This numeric directive can be modified using the ordinal : modifier. However, since it is not padded, the no-padding - and space-padding _ modifiers are not applicable to this directive.
The week number, starting from the first Sunday in the year. Days of the year that come before the first Sunday are considered to be in week zero. The number is padded with zeroes if it is fewer than two digits long. The week number is always in the range 00 through 53.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
The weekday number, starting with Sunday and ending with Saturday. The number is always in the range 0 through 6.
Note that you will get the same results parsing a weekday number with the %w directive regardless of whether the timestamp was written with the %u or the %w weekday number directive.
This numeric directive can be modified using the ordinal : modifier. However, since it is not padded, the no-padding - and space-padding _ modifiers are not applicable to this directive.
The week number, starting from the first Monday in the year. Days of the year that come before the first Monday are considered to be in week zero. The number is padded with zeroes if it is fewer than two digits long. The week number is always in the range 00 through 53.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
This directive is rewritten as %m/%d/%y.
This directive is rewritten as %H:%M:%S.
Two-digit calendar year number, e.g. 18 when the year is 2018. The number is padded with zeroes if it is fewer than two digits long.
When parsing a two-digit year number in a timestamp that does not also have a century number %C, numbers less than or equal to 68 are considered to belong to the 21st century (e.g. 2068) whereas numbers greater than 68 are considered to belong to the 20th century (e.g. 1969). To avoid the possibility of reading the wrong year number from a timestamp outside the range 1969-2068, it is better to always write timestamps using the full calendar year directive %Y instead of the two-digit year %y.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
Full calendar year number. The number is padded with zeroes if it is fewer than four digits long. Note that it's possible for the year number to be negative.
Like other numeric directives, this one can be modified using the no-padding -, space-padding _, and ordinal : modifiers.
Negative year numbers are also padded to four digits, e.g. -0099 when the calendar year number was -99 and the number was zero-padded or -99 when the number was space-padded.
The full calendar year can also be made unsigned using the ^ modifier, i.e. %^Y. This is intended to be used in combination with the era directive %#, such that a calendar year of 0 produces outputs like 1 BCE and a calendar year of -99 produces outputs like 100 BCE. Note that this seemingly off-by-one scheme conforms to the ISO 8601 standard, in which year zero represents 1 BCE.
Timezone offset in hours and minutes. The offset is represented as a sign followed by two hour digits and then two minute digits. The hour and minute digits may optionally be delimited by a colon ":". For example, different offsets might be written as +0130 or -02:00. The strptime function also recognizes the sign "±" for offset 00:00, but strftime will never write this character.
The : modifier indicates whether the timezone offset should be formatted with a delimiter or not. When passing a format string to strftime, the modified %:z directive would write +01:00 where the unmodified %z directive would write +0100.
Timezone name, or offset in hours and minutes. The offset is represented in the same way as with the timezone offset %z directive. However, when parsing a timestamp, abbreviations such as UTC, EEST, or EDT will also be recognized. The complete list of recognized abbreviations and their associated offsets can be found in the timezone-names.js source file.
Note that IANA timezone names such as America/Los_Angeles or Europe/Paris are not currently recognized, although this may change in the future.
When writing a format with strftime which includes the %Z directive, the only timezone name that is written is "UTC" when the offset is zero. Otherwise, the offset is written in the same manner as the timezone offset %z directive.
The : modifier indicates whether the timezone offset should be formatted with a delimiter or not. When passing a format string to strftime, the modified %:Z directive would write +01:00 where the unmodified %Z directive would write +0100.
This directive is rewritten as %a %b %e %H:%M:%S %Z %Y.
Era name, i.e. "CE" or "BCE" with the default English language settings. This should normally be used in combination with the unsigned calendar year directive %^Y
The "CE" and "BCE" strings can be changed by passing an options object to strftime or strptime which includes an eraNames attribute. The value of this attribute must be an array two elements, a "CE" equivalent and then a "BCE" equivalent.
The ^ modifier can be used to write switch-cased outputs, e.g. "ce" and "bce" with the default English language settings.
The %# directive, when used in a format for strptime, is case-insensitive. This means that, for example, the string "cE" will be correctly parsed no matter whether the parsing directive is modified or not.
This directive literally writes the percent character % when formatting a date with strftime and matches the same percent character when parsing a timestamp using strptime.