Importing a EBCDIC File

Hi Aveen,

Did you try importing it through the text option and what happened?

Brian

Hi Brian
I did. The data is all fuzzy.
regards
Aveeb

That is strange.  Hard for me to make a suggestion without actually seeing the data.

Hi Aveen,

You will have to use the Advance Record Definition Editor to do this.  So in the import assistant select the Record Definition Editor and your file, you should see something like this (I am on V8 right now :-( )

From this I tried to line things up a bit.  By playing with the file it seems to have a header lenght of 4 and a record lenght of 87.

As the file doesn't quite line up as you go down I next went to try and flatten it so that it would be the same width.

You will then have to select the Delimiter, I selected Carriage return/line feed and that seemed to work ok.

You will then have the file flatten dialog.  For Flatten Specifications I selected the IBM VLR ANSI format with a record lenght of 87.

I then saved the flattened file and was able to define the fields.

I selected save as to save the definition and then Exit at which time I was given the option to import the file.  If I wanted to do the import later I would then select Text for the import, select the EBCDIC file and select the record definition.

Hopefully this helps you out, not the easiest way of doing and and it took me a few trial and errors to get it right.  Been a long time since I played with EBCDIC files.

Hi Brian
Thank you very much. I'll give it a go.
Regards
Aveen

When you see a bunch of @@@@ in an ASCII editor, it would clearly be an EBCDIC file. EBCDIC files comes from IBM mainframes so normally you can obtain the record layout information from a data dictionary.

If you have this information, you can merely enter the start and len for each field and of course you would also know that entire record length maximum.

Probably the most common computer that still outputs EBCDIC  files is the AS400.

Hi, is it possible to automate the flatten file?