Merge Sequential Files - handling missing files

itgirlinuk · Post by **itgirlinuk** » Mon Sep 07, 2009 10:30 am

Hi,

I need to merge 5 different sequential files. The 5 files have the same format.

I have created a server job with a Link Collector and the 5 sequential files form as input. The merged ouput needs to be written to another sequential file.

The job works fine if all the 5 files exist in the directory. If one of them does not exist (which is entirely possible and acceptable), the job fails.

Is there a way around this? I don't want the job to fail. I want it to merge existing files and ignore missing files. Can someone tell me how to do this?

Thanks

ArndW · Post by **ArndW** » Mon Sep 07, 2009 10:53 am

Can you just add a before-job UNIX call to "touch" all the files, this would create an empty file for the ones that don't exist and just update the timestamp on those that do exist.

DSguru2B · Post by **DSguru2B** » Mon Sep 07, 2009 3:59 pm

You will be need some type of mechanism to identify the missing files and then using the touch command as described by ArndW.

chulett · Post by **chulett** » Mon Sep 07, 2009 4:54 pm

In other words, you cannot 'ignore' the missing ones like that. How about a simple 'cat' instead using wildcards? It won't care.

dxk9 · Post by **dxk9** » Mon Sep 07, 2009 10:40 pm

Howa about a wait for file stage in front of each input file stage. You can specify in the wait for file stage whether to wait for the file or continue regardless of whether the file exists or not.

Regards,
Divya

roy · Post by **roy** » Tue Sep 08, 2009 10:24 am

Hi,
You can use a routine to build a TYPE command with the existing files and run that as a filter command and use it as the merged file or write it to a target file if that is what you need.
IHTH (I Hope This Helps),

ray.wurlod · Post by **ray.wurlod** » Tue Sep 08, 2009 3:13 pm

It's a cat command on Unix, type command on Windows.

roy · Post by **roy** » Thu Sep 10, 2009 3:06 pm

Indeed it is Ray
Must have missed that unix for OS