Validate the datetime value: Difference between revisions

From LemonWiki共筆
Jump to navigation Jump to search
mNo edit summary
Line 5: Line 5:
== MySQL approach ==
== MySQL approach ==


=== Verify the value is not 0000-00-00 00:00:00 ===
=== Verify the column value is a valid (non-zero) datetime e.g. 0000-00-00 00:00:00 ===
Schema of column value
Schema of column value


Line 15: Line 15:
FROM `my_table`
FROM `my_table`
WHERE `my_date_column` LIKE '1900-%';
WHERE `my_date_column` LIKE '1900-%';
</pre>
=== Verify the date_column is a valid date e.g. YYYY-MM-DD ===
Schema of column date_column
* Type: {{kbd | key=varchar}} or {{kbd | key=date}}
* Expected format: YYYY-MM-DD
<pre>
SELECT CAST(date_column AS DATE) AS date_column
FROM my_table
WHERE CAST(date_column AS DATE) IS NULL
  AND date_column IS NOT NULL;
</pre>
</pre>



Revision as of 12:19, 6 May 2026

Validate the value should be well-formated value e.g. year-month-day 2026-05-06

MySQL approach

Verify the column value is a valid (non-zero) datetime e.g. 0000-00-00 00:00:00

Schema of column value

  • Type: datetime
  • Default value: 0000-00-00 00:00:00 (NOT allow be NULL)、"1999/12/31 上午 12:00:00" on Google sheet
SELECT `my_date_column`
FROM `my_table`
WHERE `my_date_column` LIKE '1900-%';


Verify the date_column is a valid date e.g. YYYY-MM-DD

Schema of column date_column

  • Type: varchar or date
  • Expected format: YYYY-MM-DD
SELECT CAST(date_column AS DATE) AS date_column
FROM my_table
WHERE CAST(date_column AS DATE) IS NULL
  AND date_column IS NOT NULL;

Verify the year of value is not 1900 or 1999 years

Schema of column value

  • Type: VARCHAR
  • Default value: allow be NULL

Abnormal values e.g. 1900-01-08 00:00:00 because the Excel datetime issue[1].

SELECT `my_date_column`
FROM `my_table`
WHERE `my_date_column` LIKE '1900-%';

Or list the year list

SELECT 
DISTINCT LEFT(`my_date_column`, 4)
FROM `my_table`
ORDER BY LEFT(`my_date_column`, 4)
LIMIT 50

Verify the value is null

Verify the value of DAY()[2] is NULL

Schema of column value

  • Type: datetime or VARCHAR
  • Default value: allow be NULL
SELECT `my_date_column`
FROM `my_table`
WHERE DAY(`my_date_column`) IS NULL;
SELECT `my_date_column`, STR_TO_DATE(`my_date_column`, '%Y-%m-%d')
FROM `my_table`
WHERE
STR_TO_DATE(`my_date_column`, '%Y-%m-%d') IS NULL;

Verify the value should be year/month/day format

SELECT `my_date_column`, STR_TO_DATE(`my_date_column`, '%Y/%m/%d')
FROM `my_table`
WHERE
STR_TO_DATE(`my_date_column`, '%Y/%m/%d') IS NULL;

Verify the value should be hour:minute:second format

e.g. 13:45:06

SELECT `my_time_column`, STR_TO_DATE(`my_time_column`, '%H:%i:%S') 
FROM `my_table`
WHERE
STR_TO_DATE(`my_time_column`, '%H:%i:%S') IS NULL;

Verify the value should be year-month-day hour:minute:second format

e.g. 2026-05-06 13:45:06

SELECT `my_time_column`, STR_TO_DATE(`my_time_column`, '%Y-%m-%d %H:%i:%S') 
FROM `my_table`
WHERE
STR_TO_DATE(`my_time_column`, '%Y-%m-%d %H:%i:%S') IS NULL;


Verify the value not start with four number, which means year

Icon_exclaim.gif Not recommended!

SELECT `my_date_column`
FROM `my_table`
WHERE `my_date_column` NOT REGEXP '^[0-9]{4}';

PHP appracoh

Using PHP strtotime function

Excel approach

  1. Validate the total length of cell value is equal to 10 e.g. =IF(LEN(B2)<>10, TRUE, FALSE)
  2. Validate the positions of the - symbols
  3. Combined the above condition, the verify formula is: (If the cell value is well-formatted, the formula returns true)
=NOT(
     IF(OR(
           ISERROR(FIND("-", B2, 1)), 
           ISERROR(FIND("-", B2, 6)), 
           IF(LEN(B2)<>10, TRUE, FALSE)
           ), TRUE, FALSE)
 )

References