ZetaSQL supports the following date functions.
CURRENT_DATE()
CURRENT_DATE(time_zone_expression)
CURRENT_DATE
Description
Returns the current date.
This function supports the following arguments:
time_zone_expression
: ASTRING
expression that represents a time zone. If no time zone is specified, the default time zone, which is implementation defined, is used. If this expression is used and it evaluates toNULL
, this function returnsNULL
.
Return Data Type
DATE
Examples
The following query produces the current date in the default time zone:
SELECT CURRENT_DATE() AS the_date;
+--------------+
| the_date |
+--------------+
| 2016-12-25 |
+--------------+
The following queries produce the current date in a specified time zone:
SELECT CURRENT_DATE('America/Los_Angeles') AS the_date;
+--------------+
| the_date |
+--------------+
| 2016-12-25 |
+--------------+
SELECT CURRENT_DATE('-08') AS the_date;
+--------------+
| the_date |
+--------------+
| 2016-12-25 |
+--------------+
The following query produces the current date in the default time zone. Parentheses are not needed if the function has no arguments.
SELECT CURRENT_DATE AS the_date;
+--------------+
| the_date |
+--------------+
| 2016-12-25 |
+--------------+
When a column named current_date
is present, the column name and the function
call without parentheses are ambiguous. To ensure the function call, add
parentheses; to ensure the column name, qualify it with its
range variable. For example, the
following query will select the function in the the_date
column and the table
column in the current_date
column.
WITH t AS (SELECT 'column value' AS `current_date`)
SELECT current_date() AS the_date, t.current_date FROM t;
+------------+--------------+
| the_date | current_date |
+------------+--------------+
| 2016-12-25 | column value |
+------------+--------------+
EXTRACT(part FROM date_expression)
Description
Returns the value corresponding to the specified date part. The part
must
be one of:
DAYOFWEEK
: Returns values in the range [1,7] with Sunday as the first day of the week.DAY
DAYOFYEAR
WEEK
: Returns the week number of the date in the range [0, 53]. Weeks begin with Sunday, and dates prior to the first Sunday of the year are in week 0.WEEK(<WEEKDAY>)
: Returns the week number of the date in the range [0, 53]. Weeks begin onWEEKDAY
. Dates prior to the firstWEEKDAY
of the year are in week 0. Valid values forWEEKDAY
areSUNDAY
,MONDAY
,TUESDAY
,WEDNESDAY
,THURSDAY
,FRIDAY
, andSATURDAY
.ISOWEEK
: Returns the ISO 8601 week number of thedate_expression
.ISOWEEK
s begin on Monday. Return values are in the range [1, 53]. The firstISOWEEK
of each ISO year begins on the Monday before the first Thursday of the Gregorian calendar year.MONTH
QUARTER
: Returns values in the range [1,4].YEAR
ISOYEAR
: Returns the ISO 8601 week-numbering year, which is the Gregorian calendar year containing the Thursday of the week to whichdate_expression
belongs.
Return Data Type
INT64
Examples
In the following example, EXTRACT
returns a value corresponding to the DAY
date part.
SELECT EXTRACT(DAY FROM DATE '2013-12-25') AS the_day;
+---------+
| the_day |
+---------+
| 25 |
+---------+
In the following example, EXTRACT
returns values corresponding to different
date parts from a column of dates near the end of the year.
SELECT
date,
EXTRACT(ISOYEAR FROM date) AS isoyear,
EXTRACT(ISOWEEK FROM date) AS isoweek,
EXTRACT(YEAR FROM date) AS year,
EXTRACT(WEEK FROM date) AS week
FROM UNNEST(GENERATE_DATE_ARRAY('2015-12-23', '2016-01-09')) AS date
ORDER BY date;
+------------+---------+---------+------+------+
| date | isoyear | isoweek | year | week |
+------------+---------+---------+------+------+
| 2015-12-23 | 2015 | 52 | 2015 | 51 |
| 2015-12-24 | 2015 | 52 | 2015 | 51 |
| 2015-12-25 | 2015 | 52 | 2015 | 51 |
| 2015-12-26 | 2015 | 52 | 2015 | 51 |
| 2015-12-27 | 2015 | 52 | 2015 | 52 |
| 2015-12-28 | 2015 | 53 | 2015 | 52 |
| 2015-12-29 | 2015 | 53 | 2015 | 52 |
| 2015-12-30 | 2015 | 53 | 2015 | 52 |
| 2015-12-31 | 2015 | 53 | 2015 | 52 |
| 2016-01-01 | 2015 | 53 | 2016 | 0 |
| 2016-01-02 | 2015 | 53 | 2016 | 0 |
| 2016-01-03 | 2015 | 53 | 2016 | 1 |
| 2016-01-04 | 2016 | 1 | 2016 | 1 |
| 2016-01-05 | 2016 | 1 | 2016 | 1 |
| 2016-01-06 | 2016 | 1 | 2016 | 1 |
| 2016-01-07 | 2016 | 1 | 2016 | 1 |
| 2016-01-08 | 2016 | 1 | 2016 | 1 |
| 2016-01-09 | 2016 | 1 | 2016 | 1 |
+------------+---------+---------+------+------+
In the following example, date_expression
falls on a Sunday. EXTRACT
calculates the first column using weeks that begin on Sunday, and it calculates
the second column using weeks that begin on Monday.
WITH table AS (SELECT DATE('2017-11-05') AS date)
SELECT
date,
EXTRACT(WEEK(SUNDAY) FROM date) AS week_sunday,
EXTRACT(WEEK(MONDAY) FROM date) AS week_monday FROM table;
+------------+-------------+-------------+
| date | week_sunday | week_monday |
+------------+-------------+-------------+
| 2017-11-05 | 45 | 44 |
+------------+-------------+-------------+
DATE(year, month, day)
DATE(timestamp_expression)
DATE(timestamp_expression, time_zone_expression)
DATE(datetime_expression)
Description
Constructs or extracts a date.
This function supports the following arguments:
year
: TheINT64
value for year.month
: TheINT64
value for month.day
: TheINT64
value for day.timestamp_expression
: ATIMESTAMP
expression that contains the date.time_zone_expression
: ASTRING
expression that represents a time zone. If no time zone is specified withtimestamp_expression
, the default time zone, which is implementation defined, is used.datetime_expression
: ADATETIME
expression that contains the date.
Return Data Type
DATE
Example
SELECT
DATE(2016, 12, 25) AS date_ymd,
DATE(DATETIME '2016-12-25 23:59:59') AS date_dt,
DATE(TIMESTAMP '2016-12-25 05:30:00+07', 'America/Los_Angeles') AS date_tstz;
+------------+------------+------------+
| date_ymd | date_dt | date_tstz |
+------------+------------+------------+
| 2016-12-25 | 2016-12-25 | 2016-12-24 |
+------------+------------+------------+
DATE_ADD(date_expression, INTERVAL int64_expression date_part)
Description
Adds a specified time interval to a DATE.
DATE_ADD
supports the following date_part
values:
DAY
WEEK
. Equivalent to 7DAY
s.MONTH
QUARTER
YEAR
Special handling is required for MONTH, QUARTER, and YEAR parts when the date is at (or near) the last day of the month. If the resulting month has fewer days than the original date's day, then the resulting date is the last date of that month.
Return Data Type
DATE
Example
SELECT DATE_ADD(DATE '2008-12-25', INTERVAL 5 DAY) AS five_days_later;
+--------------------+
| five_days_later |
+--------------------+
| 2008-12-30 |
+--------------------+
DATE_SUB(date_expression, INTERVAL int64_expression date_part)
Description
Subtracts a specified time interval from a DATE.
DATE_SUB
supports the following date_part
values:
DAY
WEEK
. Equivalent to 7DAY
s.MONTH
QUARTER
YEAR
Special handling is required for MONTH, QUARTER, and YEAR parts when the date is at (or near) the last day of the month. If the resulting month has fewer days than the original date's day, then the resulting date is the last date of that month.
Return Data Type
DATE
Example
SELECT DATE_SUB(DATE '2008-12-25', INTERVAL 5 DAY) AS five_days_ago;
+---------------+
| five_days_ago |
+---------------+
| 2008-12-20 |
+---------------+
DATE_DIFF(date_expression_a, date_expression_b, date_part)
Description
Returns the whole number of specified date_part
intervals between two
DATE
objects (date_expression_a
- date_expression_b
).
If the first DATE
is earlier than the second one,
the output is negative.
DATE_DIFF
supports the following date_part
values:
DAY
WEEK
This date part begins on Sunday.WEEK(<WEEKDAY>)
: This date part begins onWEEKDAY
. Valid values forWEEKDAY
areSUNDAY
,MONDAY
,TUESDAY
,WEDNESDAY
,THURSDAY
,FRIDAY
, andSATURDAY
.ISOWEEK
: Uses ISO 8601 week boundaries. ISO weeks begin on Monday.MONTH
, except when the first two arguments areTIMESTAMP
objects.QUARTER
YEAR
ISOYEAR
: Uses the ISO 8601 week-numbering year boundary. The ISO year boundary is the Monday of the first week whose Thursday belongs to the corresponding Gregorian calendar year.
Return Data Type
INT64
Example
SELECT DATE_DIFF(DATE '2010-07-07', DATE '2008-12-25', DAY) AS days_diff;
+-----------+
| days_diff |
+-----------+
| 559 |
+-----------+
SELECT
DATE_DIFF(DATE '2017-10-15', DATE '2017-10-14', DAY) AS days_diff,
DATE_DIFF(DATE '2017-10-15', DATE '2017-10-14', WEEK) AS weeks_diff;
+-----------+------------+
| days_diff | weeks_diff |
+-----------+------------+
| 1 | 1 |
+-----------+------------+
The example above shows the result of DATE_DIFF
for two days in succession.
DATE_DIFF
with the date part WEEK
returns 1 because DATE_DIFF
counts the
number of date part boundaries in this range of dates. Each WEEK
begins on
Sunday, so there is one date part boundary between Saturday, 2017-10-14
and Sunday, 2017-10-15.
The following example shows the result of DATE_DIFF
for two dates in different
years. DATE_DIFF
with the date part YEAR
returns 3 because it counts the
number of Gregorian calendar year boundaries between the two dates. DATE_DIFF
with the date part ISOYEAR
returns 2 because the second date belongs to the
ISO year 2015. The first Thursday of the 2015 calendar year was 2015-01-01, so
the ISO year 2015 begins on the preceding Monday, 2014-12-29.
SELECT
DATE_DIFF('2017-12-30', '2014-12-30', YEAR) AS year_diff,
DATE_DIFF('2017-12-30', '2014-12-30', ISOYEAR) AS isoyear_diff;
+-----------+--------------+
| year_diff | isoyear_diff |
+-----------+--------------+
| 3 | 2 |
+-----------+--------------+
The following example shows the result of DATE_DIFF
for two days in
succession. The first date falls on a Monday and the second date falls on a
Sunday. DATE_DIFF
with the date part WEEK
returns 0 because this date part
uses weeks that begin on Sunday. DATE_DIFF
with the date part WEEK(MONDAY)
returns 1. DATE_DIFF
with the date part ISOWEEK
also returns 1 because
ISO weeks begin on Monday.
SELECT
DATE_DIFF('2017-12-18', '2017-12-17', WEEK) AS week_diff,
DATE_DIFF('2017-12-18', '2017-12-17', WEEK(MONDAY)) AS week_weekday_diff,
DATE_DIFF('2017-12-18', '2017-12-17', ISOWEEK) AS isoweek_diff;
+-----------+-------------------+--------------+
| week_diff | week_weekday_diff | isoweek_diff |
+-----------+-------------------+--------------+
| 0 | 1 | 1 |
+-----------+-------------------+--------------+
DATE_TRUNC(date_expression, date_part)
Description
Truncates a DATE
value to the granularity of date_part
. The DATE
value
is always rounded to the beginning of date_part
, which can be one of the
following:
DAY
: The day in the Gregorian calendar year that contains theDATE
value.WEEK
: The first day of the week in the week that contains theDATE
value. Weeks begin on Sundays.WEEK
is equivalent toWEEK(SUNDAY)
.WEEK(WEEKDAY)
: The first day of the week in the week that contains theDATE
value. Weeks begin onWEEKDAY
.WEEKDAY
must be one of the following:SUNDAY
,MONDAY
,TUESDAY
,WEDNESDAY
,THURSDAY
,FRIDAY
, orSATURDAY
.ISOWEEK
: The first day of the ISO 8601 week in the ISO week that contains theDATE
value. The ISO week begins on Monday. The first ISO week of each ISO year contains the first Thursday of the corresponding Gregorian calendar year.MONTH
: The first day of the month in the month that contains theDATE
value.QUARTER
: The first day of the quarter in the quarter that contains theDATE
value.YEAR
: The first day of the year in the year that contains theDATE
value.ISOYEAR
: The first day of the ISO 8601 week-numbering year in the ISO year that contains theDATE
value. The ISO year is the Monday of the first week whose Thursday belongs to the corresponding Gregorian calendar year.
Return Data Type
DATE
Examples
SELECT DATE_TRUNC(DATE '2008-12-25', MONTH) AS month;
+------------+
| month |
+------------+
| 2008-12-01 |
+------------+
In the following example, the original date falls on a Sunday. Because
the date_part
is WEEK(MONDAY)
, DATE_TRUNC
returns the DATE
for the
preceding Monday.
SELECT date AS original, DATE_TRUNC(date, WEEK(MONDAY)) AS truncated
FROM (SELECT DATE('2017-11-05') AS date);
+------------+------------+
| original | truncated |
+------------+------------+
| 2017-11-05 | 2017-10-30 |
+------------+------------+
In the following example, the original date_expression
is in the Gregorian
calendar year 2015. However, DATE_TRUNC
with the ISOYEAR
date part
truncates the date_expression
to the beginning of the ISO year, not the
Gregorian calendar year. The first Thursday of the 2015 calendar year was
2015-01-01, so the ISO year 2015 begins on the preceding Monday, 2014-12-29.
Therefore the ISO year boundary preceding the date_expression
2015-06-15 is
2014-12-29.
SELECT
DATE_TRUNC('2015-06-15', ISOYEAR) AS isoyear_boundary,
EXTRACT(ISOYEAR FROM DATE '2015-06-15') AS isoyear_number;
+------------------+----------------+
| isoyear_boundary | isoyear_number |
+------------------+----------------+
| 2014-12-29 | 2015 |
+------------------+----------------+
DATE_FROM_UNIX_DATE(int64_expression)
Description
Interprets int64_expression
as the number of days since 1970-01-01.
Return Data Type
DATE
Example
SELECT DATE_FROM_UNIX_DATE(14238) AS date_from_epoch;
+-----------------+
| date_from_epoch |
+-----------------+
| 2008-12-25 |
+-----------------+
FORMAT_DATE(format_string, date_expr)
Description
Formats the date_expr
according to the specified format_string
.
See Supported Format Elements For DATE for a list of format elements that this function supports.
Return Data Type
STRING
Examples
SELECT FORMAT_DATE('%x', DATE '2008-12-25') AS US_format;
+------------+
| US_format |
+------------+
| 12/25/08 |
+------------+
SELECT FORMAT_DATE('%b-%d-%Y', DATE '2008-12-25') AS formatted;
+-------------+
| formatted |
+-------------+
| Dec-25-2008 |
+-------------+
SELECT FORMAT_DATE('%b %Y', DATE '2008-12-25') AS formatted;
+-------------+
| formatted |
+-------------+
| Dec 2008 |
+-------------+
LAST_DAY(date_expression[, date_part])
Description
Returns the last day from a date expression. This is commonly used to return the last day of the month.
You can optionally specify the date part for which the last day is returned.
If this parameter is not used, the default value is MONTH
.
LAST_DAY
supports the following values for date_part
:
YEAR
QUARTER
MONTH
WEEK
. Equivalent to 7DAY
s.WEEK(<WEEKDAY>)
.<WEEKDAY>
represents the starting day of the week. Valid values areSUNDAY
,MONDAY
,TUESDAY
,WEDNESDAY
,THURSDAY
,FRIDAY
, andSATURDAY
.ISOWEEK
. Uses ISO 8601 week boundaries. ISO weeks begin on Monday.ISOYEAR
. Uses the ISO 8601 week-numbering year boundary. The ISO year boundary is the Monday of the first week whose Thursday belongs to the corresponding Gregorian calendar year.
Return Data Type
DATE
Example
These both return the last day of the month:
SELECT LAST_DAY(DATE '2008-11-25', MONTH) AS last_day
+------------+
| last_day |
+------------+
| 2008-11-30 |
+------------+
SELECT LAST_DAY(DATE '2008-11-25') AS last_day
+------------+
| last_day |
+------------+
| 2008-11-30 |
+------------+
This returns the last day of the year:
SELECT LAST_DAY(DATE '2008-11-25', YEAR) AS last_day
+------------+
| last_day |
+------------+
| 2008-12-31 |
+------------+
This returns the last day of the week for a week that starts on a Sunday:
SELECT LAST_DAY(DATE '2008-11-10', WEEK(SUNDAY)) AS last_day
+------------+
| last_day |
+------------+
| 2008-11-15 |
+------------+
This returns the last day of the week for a week that starts on a Monday:
SELECT LAST_DAY(DATE '2008-11-10', WEEK(MONDAY)) AS last_day
+------------+
| last_day |
+------------+
| 2008-11-16 |
+------------+
PARSE_DATE(format_string, date_string)
Description
Converts a string representation of date to a
DATE
object.
format_string
contains the format elements
that define how date_string
is formatted. Each element in
date_string
must have a corresponding element in format_string
. The
location of each element in format_string
must match the location of
each element in date_string
.
-- This works because elements on both sides match.
SELECT PARSE_DATE('%A %b %e %Y', 'Thursday Dec 25 2008')
-- This doesn't work because the year element is in different locations.
SELECT PARSE_DATE('%Y %A %b %e', 'Thursday Dec 25 2008')
-- This doesn't work because one of the year elements is missing.
SELECT PARSE_DATE('%A %b %e', 'Thursday Dec 25 2008')
-- This works because %F can find all matching elements in date_string.
SELECT PARSE_DATE('%F', '2000-12-30')
When using PARSE_DATE
, keep the following in mind:
- Unspecified fields. Any unspecified field is initialized from
1970-01-01
. - Case insensitivity. Names, such as
Monday
,February
, and so on, are case insensitive. - Whitespace. One or more consecutive white spaces in the format string matches zero or more consecutive white spaces in the date string. In addition, leading and trailing white spaces in the date string are always allowed -- even if they are not in the format string.
- Format precedence. When two (or more) format elements have overlapping
information (for example both
%F
and%Y
affect the year), the last one generally overrides any earlier ones.
Return Data Type
DATE
Examples
This example converts a MM/DD/YY
formatted string to a DATE
object:
SELECT PARSE_DATE('%x', '12/25/08') AS parsed;
+------------+
| parsed |
+------------+
| 2008-12-25 |
+------------+
This example converts a YYYYMMDD
formatted string to a DATE
object:
SELECT PARSE_DATE('%Y%m%d', '20081225') AS parsed;
+------------+
| parsed |
+------------+
| 2008-12-25 |
+------------+
UNIX_DATE(date_expression)
Description
Returns the number of days since 1970-01-01.
Return Data Type
INT64
Example
SELECT UNIX_DATE(DATE '2008-12-25') AS days_from_epoch;
+-----------------+
| days_from_epoch |
+-----------------+
| 14238 |
+-----------------+