SQL(phpMyAdmin): Change Column Format of Date With Command - mysql

Question

i have a column with name 'expdate' with type 'varchar' with values like this:
2021-02-27
28-02-2023
29/02/2024
.
.
.
now i want change all date values like format %d-%m-%Y to the %Y-%m-%d and save to new column with type date and use:
UPDATE `Users`
SET `fixed_expdate` = STR_TO_DATE(REPLACE(`expdate`,"/",'-'), "%d-%m-%Y")
where `expdate` != '0000-00-00'
but sql command show error like:
SQL query:
UPDATE `Users` SET `fixed_expdate` = STR_TO_DATE(REPLACE(`expdate`,"/",'-'), "%d-%m-%Y") where `expdate` != '0000-00-00'
MySQL said: Documentation
#1411 - Incorrect datetime value: '2021-02-27' for function str_to_date
i need to sql command,What is your solution?

Answer 1

You must use different converting expressions for each separate source data format. And you must check that only the values which matches the pattern used in current query are included into the convertion.
This can be performed using a lot of separate queries, each converts one of existing source formats.
-- convert the dates which have correct format
UPDATE `Users`
SET `fixed_expdate` = `expdate`
WHERE `expdate` REGEXP '\d\d\d\d-\d\d-\d\d';
-- convert dated which looks like 28-02-2023
UPDATE `Users`
SET `fixed_expdate` = STR_TO_DATE(`expdate`, "%d-%m-%Y")
WHERE `expdate` REGEXP '\d\d-\d\d-\d\d\d\d';
-- convert dated which looks like 29/02/2024
UPDATE `Users`
SET `fixed_expdate` = STR_TO_DATE(`expdate`, "%d\/%m\/%Y")
WHERE `expdate` REGEXP '\d\d\/\d\d\/\d\d\d\d';
-- and so on
After each separate convertion (or after a lot of convertions) you may look at the rows which were not converted
SELECT `expdate`
FROM `Users`
WHERE `fixed_expdate` IS NULL
LIMIT XXX
and build the next converting expression for a pattern which was not used yet - until all values are successfully converted.
MySQL said: Documentation
#1411 - Incorrect datetime value: '2021-02-27' for function str_to_date
This is due to 2 separate formats which differs in day and month posession (which are swapped) - they cannot be distinguished easily. You must use additional checking if both formats are present in the data. And you cannot distinguish what format must be applied if both formats are possible (for example, when the value is '01/02/2021').
UPDATE (copied from the comment)
Your regular expressions should should probably have anchors for the beginning and end of the string ('^' and '$'). – Gordon Linoff