DSXchange

I suspect (from my understanding of the metadata model) that justification cannot be specified in metadata, but am willing to be corrected. Metadata are far more about how things are stored, rather than how they're displayed. In general.

There's no explicit mention of an eleventh row in your original question, which is probably why it was not addressed in my answer. However, you could create a second output link that outputs only the negative of the total when the rowcount is 10. Use any appropriate stage type downstream to combine ...

Might be a good candidate for using Transformer stage looping. Count from 1 to 10, initialize or increment the loop variable that's keeping the ongoing sum.

Did you try it?

You could use a hashed file. When reading back from the hashed file include a sorting clause (for example BY fieldname or BY.DSND fieldname) in the Selection field.

/.dshome is not used for itag installations. The DSHOME environment variable is used.

Please encase your examples in Code tags so that we can discern what goes with what.

I make great use of the ability to drag a full or partial table definition onto a link.

DataStage Designer recognises single click, double click, triple click, quatruple click and, in one case that I've discovered, even quintuple click.

The check box to which Stuart refers appears in the dialog where you set up and submit the job run for Column Analysis. By default it is unchecked to save resources that are needed to compare large strings.

Have you registered interest in the table metadata and validated settings and connections within the project in which you're trying to run Column Analysis?

Maybe the simplest solution is to use ODBC to connect to Oracle.

You need to enable "reference link with multi-row result set".

Right click in the background of the Transformer stage and open Stage Properties. Choose the Inputs tab, then the reference input from the drop down list of link names. The check box you need is at the bottom of the page.

ssnegi wrote:you can use merge stage. first file - source file (one record) . Second file - reference file (10K records). Join Type - Right Outer Join. You can use transformer for any tranformation before or after the merge stage.

Not true.
The server Merge stage performs SQL joins on a pair of text files.

Also check for data type changes on the way through (including perhaps ustring to string or vice versa).