I have a column named StartDate
containing a date in this format: 03-03-2012 15:22
What I need is to convert it to date. It should be looking like this: DD/MM/YYYY
What I have tried without success is:
select p1.PA_VALUE as StartDate, p2.PA_VALUE as EndDate from WP_Work p LEFT JOIN PARAMETER p1 on p1.WP_ID=p.WP_ID AND p1.NAME = 'StartDate' LEFT JOIN PARAMETER p2 on p2.WP_ID=p.WP_ID AND p2.NAME = 'Date_To' WHERE p.TYPE = 'EventManagement2' AND TO_DATE(p1.PA_VALUE, 'DD/MM/YYYY') >= TO_DATE('25/10/2012', 'DD/MM/YYYY') AND TO_DATE(p2.PA_VALUE, 'DD/MM/YYYY') <= TO_DATE('26/10/2012', 'DD/MM/YYYY')
Is there a way to do this?
EDIT1: the PA_VALUE
column is: VARCHAR2
Advertisement
Answer
When you convert your string to a date you need to match the date mask to the format in the string. This includes a time element, which you need to remove with truncation:
select p1.PA_VALUE as StartDate, p2.PA_VALUE as EndDate from WP_Work p LEFT JOIN PARAMETER p1 on p1.WP_ID=p.WP_ID AND p1.NAME = 'StartDate' LEFT JOIN PARAMETER p2 on p2.WP_ID=p.WP_ID AND p2.NAME = 'Date_To' WHERE p.TYPE = 'EventManagement2' AND trunc(TO_DATE(p1.PA_VALUE, 'DD-MM-YYYY HH24:MI')) >= TO_DATE('25/10/2012', 'DD/MM/YYYY') AND trunc(TO_DATE(p2.PA_VALUE, 'DD-MM-YYYY HH24:MI')) <= TO_DATE('26/10/2012', 'DD/MM/YYYY')
Outside the scope of the question, but storing dates as strings is bad practice, and storing date times is even worse.
- We need to convert the strings to dates in order to do any form of date processing (arithmetic, interval assessment, etc) on them
- Strings offer no guarantees regarding format, so we run the risk of date corruption crashing our code. We can defend against this by employing
VALIDATE_CONVERSION()
(available since 12c, find out more ) but it’s still a PITN - Using non-standard datatypes makes it harder to reason about the data model and the code we build over it.