I'm coding a batch application using JDBC. The batch app insert data it reads from a file into a Oracle table.
On error, its supposed to try to re-insert all rows from the beginning.
In the end the app will receive a lot of Ora-00001 errors, meaning that a dublicate key constraint is being violated. but ora-00001 is an Oracle specific code. and theres no way of writing database independent code using pure JDBC API.
Now I've learned that Oracle has introduced a ignore_row_on_dupkey_index hint in the 11g version.
What I want to ask is actually not technical.
Why don't they just define a set of those solutions / codes e.g. into the ANSI standard, say; SQL_Error_001 means from now on "dublicate key" violation. period.
I mean, smart people suggest these standards, like SQL-92 etc., right ?
What's behind this, is there a reason for them to leave those cases out of the standards they define ?
Wouldn't life be so much simpler if they did define common cases (like dublicate keys) into standards ?
Edit : Also, I've always wondered why they don't introduce a simple "calendar" / "datepicker" tag in HTML that would make the browser open a standard 'datepicker' instead of leaving that to the web programmer.
Last edited by msegmx; 02-04-13 at 03:47.
Reason: Ansi-92 --> SQL-92