Parsing and Format of Temporal Variables

DolphinDB provides a function temporalParse to convert a string with specified format to DolphinDB temporal variable, and function temporalFormat to convert a DolphinDB temporal variable to a string with specified format.

The following table shows the temporal formats in DolphinDB:

The parameter format in function temporalParse and temporalFormat has 2 types of representation:

• With deliminator(s)

Any symbol or character is treated as a deliminator except the characters that are used to express a temporal format: y, M, d, H, h, m, s, a, S, and n. A deliminator in the parameter format should be identical as the deliminator in the input string.

temporalParse("14-02-2018","dd-MM-yyyy");
// output: 2018.02.14

temporalParse("14-02-2018","dd/MM/yyyy");
// output: 00d

temporalParse("14//02//2018","dd//MM//yyyy");
// output: 2018.02.14

temporalParse("14//02//2018","dd/MM/yyyy");
// output: 00d

temporalParse("14//02//2018","dd..MM..yyyy");
// output: 00d

We can simplify the formats by using a single letter between deliminators for the parameter format. For example, we can use the format "y/M/d" instead of "yyyy/MM/dd" for "2018/01/16". As "y" may mean both "yyyy" and "yy", for this case the system decides on the format based on the number of digits between deliminators.

temporalParse("14-02-18","d-M-y");
// output: 2018.02.14

temporalParse("2018/2/6 02:33:01 PM","y/M/d h:m:s a");
// output: 2018.02.06T14:33:01

"MMM","SSS", "nnnnnn" and "nnnnnnnnn", however, cannot be simplified to a single letter.

temporalParse("02-FEB-2018","d-MMM-y");
// output: 2018.02.02

temporalParse("02-FEB-2018","d-M-y");
// output: 00d

temporalParse("13:30:10.001","H:m:s.SSS");
// output: 13:30:10.001

temporalParse("13:30:10.001","H:m:s.S");
// output: Invalid temporal format: 'H:m:s.S'. Millisecond (S) must have three digits.

temporalParse("13:30:10.008001","H:m:s.nnnnnn");
// output: 13:30:10.008001000

temporalParse("13:30:10.008001","H:m:s.n");
// output: Invalid temporal format: 'H:m:s.n'. Nanosecond (n) must have six or nine digits.

The temporalParse function is very flexible in interpreting the numbers between deliminators in the input string.

temporalParse("2-4-18","d-M-y");
// output: 2018.04.02

temporalParse("2-19-6","H-m-s");
// output: 02:19:06

temporalParse("002-019-006","H-m-s");
// output: 02:19:06

For millisecond, microsecond and nanosecond, however, the corresponding number of digits in the input string must be 3, 6 and 9 respectively.

temporalParse("2018/2/6 13:30:10.001","y/M/d H:m:s.SSS");
// output: 2018.02.06T13:30:10.001

temporalParse("2018/2/6 13:30:10.01","y/M/d H:m:s.SSS");
// output: 00T

temporalParse("2018/2/6 13:30:10.000001","y/M/d H:m:s.nnnnnn");
// output: 2018.02.06T13:30:10.000001000

temporalParse("2018/2/6 13:30:10.0000010","y/M/d H:m:s.nnnnnn");
// output: 00N

In comparision, for function temporalFormat, the number of characters between deliminators determines the number of digits in the output.

temporalFormat(2018.02.14,"dd-MM-yyyy");
// output: 14-02-2018

temporalFormat(2018.02.14,"dd/MMM/yy");
// output: 14/FEB/18

temporalFormat(02:19:06,"HH.mm.ss");
// output: 02.19.06

• Without deliminators

For this reprensentation, the parameter format must be composed of the formats in the temporal formats table. We cannot use a single letter to represent a format in the temporal format table.

temporalParse("20180214","yyyyMMdd");
// output: 2018.02.14

temporalParse("122506","MMddyy");
// output: 2006.12.25

temporalParse("155950","HHmmss");
// output: 15:59:50

temporalParse("035901PM","hhmmssaa");
// output: 15:59:01

temporalParse("02062018155956001000001","MMddyyyyHHmmssnnnnnnnnn");
// output: 2018.02.06T15:59:56.001000001