DSXchange

Hi, I am wondering if anyone has come across this problem before and has found away round it? Here is the multivalued xml that I am presented with from the source system: <c20>Y</c20> <c20 m='12'>Y</c20> <c20 m='14'>Sole</c20> It should really be like this, but isn't: <c20> m='1'>Y</c20> <c20 m='12'...

Try this:

* ?="metal chassis"
COPY_A [1] {OutputColumn}

or if you want to be more specific:

A|B|?|^
COPY_A [3] {OutputColumn}

COPY_A will copy tokens containing a space

If you want to clear the value of the token afterwards add RETYPE [x] 0

Hope this helps.

Bob.

Bear in mind that NLS works at "two levels". You need to have an NLS code page enabled in the dsenv file for view data to work as well as in the project/job. If you only define NLS in the project/job the data may well get transferred correctly but you won't be able to see it unless you loa...

In the project I am working at the moment we have had to load tables with something like 300m rows and the prospect of a failure and restart could not be considered. What we did was to create a control table which contained the first part of the key to the source table records so that we could load ...

If you are running 10 iterations of the same job then check your log purge settings.

If it is set, for example to 3, then you will only see the last three iterations that are running.

Set it to a higher value or clear after 1 day and see if that helps.

Hope that helps!

Bob.

Hello, I am outputting xml data to a flat file using the xml satge. Everything is working fine except that the xml declaration on the first line of the output is always <?xml version="1.0" encoding="UTF-8"?> (On the XML_Composer Step/Header the "Include XML Declaration is ch...

Ok, many thanks
Bob.

Hello, I have noticed that DataStage is steadily filling up the disk share which handles datasets and scratch space and also the xmeta data. I have cleared down the datasets and also have a process which keeps the job logs cleared down as much as possible. However there are a couple of folders with ...

You could try multiplying the value 100 million, the DecimalToString, then substring off the characters you want.

Bob.

Yes, it was definitely something simple!

STRING Low_Bal; is fine

What was causing the problem was the assignment:

Low_Bal='aaaa' causes the compilation error messages and char behaviour

Low_Bal="aaaa" works just fine!

Thanks for the help.

Bob.

Hi Eric,

Thanks for the suggestion and the link.

Still no joy, I'm afraid!

The link obviously shows "pure C++" code, I think the build op code is a "version" of C++ requiring a slightly different approach.

Bob.

Hi Ray, I have tried all sorts of variations on a theme without success: (STRING is a type from APT_String in the declaration section of the generated C code) STRING Low_Bal (10); - same error STRING Low_Bal 10; - expected initializer before numeric constant STRING Low_Bal,10; - expected unqualified...

Hello, I am trying to write a simple build op which is working except for one problem where I cannot declare a string variable (Low_Bal) Here's the Definition: #include </IIS/IBM/InformationServer/Server/PXEngine/include/apt_util/string.h> APT_String Low_Bal Here's the compiler warning: warning: mul...

Hi, Basically, you need to add the header row to each row of data and then you can manipulate however you like. Read the file with the "First Line is Column Names" set to false. Use a transformer to recognise the header row with something like Year = "Year" and send this row as a...

We had exactly this problem.

Either run the job on one node, or partition by the columns you are using as keys for the update.

The problems is almost certainly caused by duplicate update keys.

Regards,
Bob.