Hello,
I have loaded a file with international characters from around the world in UTF-8 format using DataStage 8.0 parallel job into a Teradata Unicode table. When I export the table from BTEQ and open the file in Excel(Using UTF-8), I am able to display all the international characters.
If I run a query from my the Teradata Client tool SQL Assitant, then all the international characters appear garbled up. If I export it from the SQL assitant tool into a UTF-8 delimited file and open it up in excel, it still appears messed up. It says everywhere that SQL assitant and odbc drivers are all UTF-8 compliant.
Any direction to resolve this will be great. It works if I export from command line, but the characters get messed up if I use the client tool.
Teradata SQL Assistant - Does not display intl characters
Moderators: chulett, rschirm, roy
This is not a question for the Parallel Extender forum, it would be better suited under general forum.
I would put money down that it is actually the character set of your windows instance that is messing up your view of the data. I have seen people forget that PUTTY will display characters according to it's character mapping. Your Windows (or VDI session) needs to have that character set I believe.
You say you executed a BTEQ then opened the file up in Excel... Did you FTP the file off of Unix and onto Windows to view it? The FTP could have manipulated the endian nature of your data.
Are integers also messed up or just strings perhaps?
The Teradata SQL assistant is a windows tool. make sure that it is configured to use proper character set, don't assume it defaults to utf8.
Take a peek at the data using a hex editor and look for a knows string pattern. Dummy up an entry with "Hello World" string and look for that combination of hex characters.
I would put money down that it is actually the character set of your windows instance that is messing up your view of the data. I have seen people forget that PUTTY will display characters according to it's character mapping. Your Windows (or VDI session) needs to have that character set I believe.
You say you executed a BTEQ then opened the file up in Excel... Did you FTP the file off of Unix and onto Windows to view it? The FTP could have manipulated the endian nature of your data.
Are integers also messed up or just strings perhaps?
The Teradata SQL assistant is a windows tool. make sure that it is configured to use proper character set, don't assume it defaults to utf8.
Take a peek at the data using a hex editor and look for a knows string pattern. Dummy up an entry with "Hello World" string and look for that combination of hex characters.
Glad you sorted your issue out - but - one word of warning with sql assistant.
There is an option somewhere (preferences I guess) - that enables the use of ODBC extentions and is enabled by default.
This can cause different results when running stuff via SQL Ass vs BTEQ.
We fell over this when trying some code in DS that we had tested via SQL Ass - believe it was a DDL change and it applied the DDL incorrectly via SQL assistant.
Cant remember the details I am afraid - but one to watch out for.
There is an option somewhere (preferences I guess) - that enables the use of ODBC extentions and is enabled by default.
This can cause different results when running stuff via SQL Ass vs BTEQ.
We fell over this when trying some code in DS that we had tested via SQL Ass - believe it was a DDL change and it applied the DDL incorrectly via SQL assistant.
Cant remember the details I am afraid - but one to watch out for.