One Star

Unicode row separator in tFileInputDelimited

Hello!
I just migrated from TOS 3.x to TOS 5.0.2 and I have the following issue with the tFileInputDelimited Component:
We read CSV files in Unicode encoding and use non-standard row and field separators:
* Row separator: "\u2193"
* Field separator: "\u2192"
That was no problem in the previous versions, but no we get the following error when we try to run the Job ( and the component is marked with an red X):
* The length of Parameter (CSV Row Separator) must be smaller than "2".
Do you have any hint how to solve that issue? The file encoding is set to UTF-8.
3 REPLIES
One Star

Re: Unicode row separator in tFileInputDelimited

Hi
You can follow these steps.
1. Modify tFileInputDelimited-->Advanced Settings->Encoding. Choose "Custom" and type "Unicode".
2. Type Row separator and Filed separator with your own input method instead of "\u2193" or "\u2192".
I have tested this and it works fine.
Regards,
Pedro
One Star

Re: Unicode row separator in tFileInputDelimited

Hi
Thanks for you quick answer.
How do I define and use my own input methods?
By the way, do you have an idea where this check is done and since which version of TOS the old way to specify the row and field separator ("\u...") does not work anymore?
And: Is it possible to have different versions of the used JDK for the TOS itself and the jobs that are generated with it (Export Jobs)?
Reason for that question: We still use JDK 5 in production environment für running the jobs and TOS 5.x does not work anymore with JDK 5 (via the -vm command line argument).
Thanks and kind regards
Jürgen
One Star

Re: Unicode row separator in tFileInputDelimited

Hi
My version of TOS is 5.0.1. OS: WinXP SP3. JDK: 1.6.0_11.
And I try to reproduce this issue. Set "\u2193" or "\u2192" in Row Separator and Field Separator. But I didn't get any errors.
A little strange.
Which OS do you use?
Could you send an email to me and attach this delimited file?
Regards,
Pedro