DSXchange

The input columns from the file is taken to be of unicode . So , if some special characters come in the file , they get translated to the NLS set in the project. But in the basic transformer in the parallel job , i could not see the extended option .

Does this mean , basic transformers do not support NLS options.

I cannot check right now, but can you not set the NLS attributes for the whole BASIC transform stage in the stage properties tab? Also, the character will be translated going into the stage but also when going out, so the net effect will be transparent to you; i.e. the character might get translated entering the stage, but will get translated back to the original value when going out.

Character map translations are performed at the boundaries between DataStage and the "outside world". Within DataStage everything is managed in a UTF encoding of Unicode. Therefore (since a Transformer stage is not on any such boundary) there is no need to perform any character mapping in any kind Transformer stage.

Locales, though, are a different question, but not the one that was asked.

so , when i read from a file which has some spanish characters, and if i set the unicode option in the sequential file stage . Is that enough ?

I have a issue with a degree like symbol used in spanish along with the address .

No. As well as the Unicode option for the column, you also need to set the NLS map name used to translate from however the data are encoded in the file. This is done in the Sequential File stage, not in the Transformer stage.

my question is with the term you used as "outside world" .
So , encoding happens only when a file /dataset is read or written from disk.

File, yes. Data Set no (Data Set is "internal" to DataStage). File Set, yes. Database table, yes. FTP, yes. And so on.