in reply to Re: Re: UTF-8, Oracle and Perl life
in thread UTF-8, Oracle and Perl life

Is your current extract script either setting the vars I mentioned or inheriting them from the calling environment? for instance, if the job is running from cron, you will absolutely need to set this in the perl script. This is independent from the database setup, and your perl won't inherit the settings from Oracle. They will need to be in the calling environment (set perhaps in a calling shell script) or setting them specifically in the perl.
  • Comment on Re: Re: Re: UTF-8, Oracle and Perl life

Replies are listed 'Best First'.
Re: Re: Re: Re: UTF-8, Oracle and Perl life
by Akira71 (Scribe) on Oct 07, 2002 at 20:05 UTC
    Actually they are inheriting them from the database so no, we are not setting them in the perl script. I can add the appropriate ones to the perl script.

    I will try that and see if it helps.

    Thank you,

    Akira

[reply]
      Actually they are inheriting them from the database
      No they're not. NLS settings are not cascaded, they are set individually for each client and server. Use NLS_LANG="JAPANESE_JAPAN.UTF8" for each of your clients. Either on the command line (or in the script) for a *nix box or in the registry for MS clients.

      rdfield

[reply]
[d/l]

In Section Seekers of Perl Wisdom