DSXchange

Dummy, (I really hesitated to call you like that), I hope you don't mind that I ask Ken on the range (between) style select within this topic? Ken I had a couple of situations when I had to do such a range style lookup and couldn't help but to do a 'multi row result set' reference and subsequenly ex...

Billy

I don't understand the requirement (maybe a had one beer too much [:)]). Please try to explain it with more detail.

Stephan

Hi is the job logic and the structure of the hash file the same as originally posted? If so, you would need a secondary index on CO_CODE rather than MATERIAL_NO, PLANT_ID? Where is your stream input coming from? If it's a hash file or uv table as well, you could interchange stream and reference. Or ...

As for the cumulation as of the week before, does that mean you would always only sum it up starting Sunday (or Monday) last week up to yesterday? In most DBMS there should be a function like WEEKDAY, returning 0-6 (or 1-7) for days of the week Sunday to Saturday or Monday to Sunday. Provided that t...

Does this fit better? You would possibly have to translate null values to 0.

SELECT a.IDNo, a.Date, a.Tran amount, SUM(b.Tran amount)
from table a, outer table b
where a.IDNo=b.IDNo
and b.Date < a.Date
group by a.IDNo, a.Date, a.Tran amount


Stephan

Billy

looks like you are trying to get the maximum Date and the sum of Tran Amount per IDNo. If that's the case, couldn't you try:

select IDNo, max(Date), sum(Tran amount)
from table a
group by IDNo

[?]

Stephan

Watch for pathnames and calls to OS commands: http://www.tools4datastage.com/forum/to ... C_ID=84936

Stephan

I surely renamed a job (or created copies) in Manager more than a hundred times, but they never were corrupted. In Designer you will rather save a job under a new name, which in effect is creating a copy. Any controlling jobs refering to the original job will run the original, compiled version. Step...

Substring operations can be performed using square brackets [start,length].
As a derivation for your stage var you could use:

stagevar=sourcecolumn[4,1].


For other interesting forms to use [] please refer to the documentation.

Regards

Stephan

Not that I remember. Are you using the function Date() or the system variable @DATE? If it's the function, maybe you are not really updating your date, but re-using the value returned by the first call?

Stephan

Hello Joerg, I have no idea how the datasources are configured under Windows? Is there a configuration file or is it done in the registry? Is OCI not configured in the same place as ODBC datasources? If the datasource names were the same on each of your servers, you wouldn't need to change anything ...

If you have three different DataStage servers, you could use the same datasource names, but in the $(cat /.dshome)/.odbc.ini you would direct them to different databases.

Stephan

Cal

I think in this case

"testing.sh 'hello 1234'"

would suffice? The composition of the string like you used it is only necessary if you are using variables.


Stephan

Cal in fact I also sometimes feel the need to quote things, e.g. an email subject that contains blanks or an enumeration. It's usually helpful to first decide on which quotation mark to use on inner and outer quotation (" or '). If you want to use parameter evaluation should choose " as in...

Nitin, up to DS 5.2 you could use: ereplace(oconv(Date(), "D-YMD[4,2,2]"),'-','') ereplace(oconv(Date(), "D-YMD[2,2,2]"),'-','') I think I once tried to get the date formatted without delimiter in one step, but didn't succeed. As this does'nt seem to be possible, I eliminate it u...