LOAD with 2/3 less fields results in a bigger file!?

[datanibbler]

Hi,

I have tried to slim down an app by removing all the fields which are not further processed - to that end I removed that "LOAD * " and used the complete fieldlist instead, then commenting out all the fields which I do not want.

<=> After running that, I get a qvd file about 1/3 bigger than the original ...

Why is that?

Thanks a lot!

Best regards,

DataNibbler

[datanibbler]

Hi Peter,

I'm not doing any conversion, into dual values or otherwise, and I'm not merging any fields. I don't have any timestamps.

How do I isolate the QVD_headers to post them?

This is current again - I now have to attach a LOAD from an Excel_file to the existing SAP data, so I'd rather get rid of all the unnecessary fields first, that will make it a lot easier for myself. It seems to work fine, I'm just testing whether everything is there, but still the export from that qvw is about 20MB bigger than it used to be.

Best regards,

DataNibbler

[marcus_sommer]

Just open the qvd with an editor like notepad ++ and you will see the the xml-header and to find there the fields-part is quite easy. Then a copy+paste from there to MS Word and the same with the origin QVD and using the inbuilt comparing-feature (with only a few fields the manually comparing will be faster) and you will find each differences.

Beside this there must be changes between them caused from any actions which prevent an optimized load and/or any script-statements before the load and/or any script-statements between the load and the storing within the qvd again (like already hinted above).

- Marcus

[datanibbler]

I just have to get Notepad++ - I know that is a  handy little tool, but it's presumably not so easy on a centrally administered machine - let me see.

[marcus_sommer]

For this task the normal windows notepad should be sufficient.

[Peter_Cammaert]

Watch out. Regular Windows Notepad may choke on SAP table dumps that run in the megabytes...

[Peter_Cammaert]

You can use any text tool that is able to load your QVD and that is able to copy the header part that starts with

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>

:

and ends with

:

</QvdTableHeader>

Depending on the size of your QVD and the complexity of its creation, the XML header may consist of hundreds to thousands of text lines.