避免在Oracle SQL中假脱机时将错误复制到CSV中

Question

I have a SQL code which spools out data into a .csv file from Oracle databases.

set echo off
set feedback off
set linesize 1000
set pagesize 0
set sqlprompt ''
set trimspool on
set verify off
spool test.csv
    /*Code Part*/
/
spool off

The problem is that, if at all any error occurs (eg: resource busy issue) while executing code part, those error messages are getting copied into .csv file along with the spooled data. Is there any way to avoid it?

It would be much more helpful if some one suggests me a way to redirect these error messages to a .txt file (I dont know if it is possible or not).

Thanks!

Answer 1

You may want to look into using the UTL_FILE package to create the file you want rather than counting on SQL*Plus to redirect your output. While this would perhaps require some rewriting on your part you'd end up with better control of what is being written.

Another option would be to filter the output file to eliminate . For example, if the lines you don't want to see all start with "ORA-" you might be able to use something like the grep command:

grep -v ^"ORA-" test.csv

Or you might have to use something like sed or awk if your filtering requirements are more complicated.

Share and enjoy.