Char(0) being stored in Sequential file

[rsadevelopers]

Hi All, we have a simple job that at times writes rows with x00 to a sequential file. The job does not abort and seems to complete successfully. If we find x00 in the file we re-run the job and it works fine. Each column in the basic transformer calls a routine that removes special characters.

Job Design:

Dataset---->Filter---->Basic Transformer---->Sequential File

Here is the code for the routine that we call for each column:

Code: Select all

Ans=Ereplace(Convert(char(160)," ",Convert(char(162)," ",Convert(char(172)," ",Convert(char(194)," ",Convert(char(199)," ",Convert(char(201)," ",Convert(char(204)," ",Convert(char(233)," ",Convert(char(13)," ",Convert(char(10)," ",Convert(char(63)," ",Convert(char(0)," ",Convert("`"," ", RawString))))))))))))),'"','""')

if Index(Ans,'"',1) > 0 or Index(Ans,'|',1)>0 
then Ans = '"':trim(trim(Ans," ","L")," ","T"):'"'

Ans = trim(trim(Ans," ","L")," ","T")

Anyone know what might be going on?

Thanks in advance

[ray.wurlod]

Welcome aboard.

If you are using Char data type, then DataStage will pad these to the designated length using the character specified in APT_STRING_PADCHAR, the default value of which is 0x00.

If you want it to pad Char data types with space, change the value of APT_STRING_PADCHAR to space, or to 0x20.

[rsadevelopers]

Hi Ray, APT_STRING_PADCHAR is already set to 0x20

It's strange in that if we find nulls in the file We re-run and file look fine.

There are many jobs running at the same as the files are being created and I/O is very heavy. Not sure if this could be the cause.