Add the ability to control db2sqljcustomize behavior on untyped host variables and prevent silent data type conversion

We modernize a banking database application originally written in COBOL by migrating the functionality into a new written Java application.

During this application migration we noticed that there is an issue with unexpected/unwanted data type conversions which in return result in data loss.

E.g. given following code example will result in losing the decimal scale portion ‘0.99’ due to implicit data type conversion:

BigDecimal amountSrc = new BigDecimal(55.99);

BigDecimal amountDest = null ;;

SELECT :amountSrc*1 //-> Implicit casting to Integer

INTO :amountDest

FROM sysibm.sysdummy1

Given that ‘amountSrc‘ is an untyped input host variable and does NOT reference to any Db2 COLUMN the SQLJ customizer sends ( <? > * 1 ) for prepare to the database.

The database in return has no knowledge about the actual data type and applies Integer type as per ‘* 1’ portion of the calculation.

This implicit type conversion is a problem to our application development.

Therefore, we need a way to tell db2sqljcustomize how to handle these kind of issues. We would require an option to ensure a warning is returned during the code customization process.

Or even better have db2sqljcustomize abort the process with an error and force a developer to review their code or call db2sqljcustomize with a

new option –accept- BigDecimal-int-conversion.

Needed By

Quarter

Post comment

By clicking the "Post Comment" or "Submit Idea" button, you are agreeing to the IBM Ideas Portal Terms of Use.
Do not place IBM confidential, company confidential, or personal information into any field.

Please enter your email address

RELATED IDEAS

Add the ability to control db2sqljcustomize behavior on untyped host variables and prevent silent data type conversion