Translate age() function from postgreSQL to Google Bigquery - google-bigquery

Question

I'm aware that age() doesn't exist in Google Bigquery. There's DATE_DIFF but it's not giving me the accurate difference between 2 dates.
For example, if run this query in postgresql
select AGE('2020-10-10','2000-10-11')
This will give me the result: "19 years 11 mons 30 days"
However, if I run this query in Google Bigquery
SELECT DATE_DIFF(safe.parse_date('%Y%m%d', safe_cast(20201010 as string)),safe.parse_date('%Y%m%d', safe_cast(20001011 as string)), YEAR)
This will give me result: 20
How can I achieve the result like in postgresql? That has the difference in year, month, and day between 2 dates in Google bigquery.
Thanks.

Answer 1

Since there is no direct function available in BigQuery using which you can get the output in the expected format, you can use Javascript UDF functions which lets you call a Javascript function from a SQL query in BigQuery. For your reference, you can use the below sample query.
Query:
CREATE TEMP FUNCTION
DiffinDate(days int64)
RETURNS STRING
LANGUAGE js AS r"""
function DateDiff(days) {
var years = Math.floor(days / 365);
var months = Math.floor(days % 365 / 30);
var days = Math.floor(days % 365 % 30);
var yearsDisplay = years + (years == 1 ? " year " : " years ");
var monthsDisplay = months + (months == 1 ? " month " : " months ");
var daysDisplay = days + (days == 1 ? " day" : " days");
return yearsDisplay + monthsDisplay + daysDisplay
}
return DateDiff(days);
""";
WITH
INPUT AS (
SELECT
DATE_DIFF(DATE '2010-07-07', DATE '2008-12-25', DAY) day )
SELECT
DiffinDate(day) AS diff_in_date
FROM
INPUT
Output:
You can change the calculation logic as per your convenience.