Hi,

I'm currently migrating a SAPDB 6.2 database to 7.4 by exporting the 6.2 catalog and 
each table with xload in FORMATTED data files and importing it into a 7.4 database 
with loadercli. I almost got it working now (including LONG columns) pretty automated 
with a java tool creating 7.4 command files and fixing the data files (long column 
offsets are expressed differently without "->" in 7.4) from the 6.2 ones.

I do have a slight problem and I hope theres a simple solution to it...

For some reason some tables I want to export have hex 13 or hex 10 (CR/LF) in VARCHAR 
fields.
Yes I know this is odd...
I have no idea how this happened (its a large DB and I'm not sure what the table is 
used for).
This of course causes a line feed in the middle of a dataset if one would read the 
file line by line and not by the rows size.
loadercli gets confused and complains about missing column data in such a line, so it 
looks like loadercli loads a formatted data file line by line instead of reading just 
the rowlength (which is fixed for FORMATTED files) regardless of any line feeds.

I haven't tried to import in COMPRESSED mode, because there would be more 
modifications necessary on the 6.2 compressed data files (format for FIXED fields is 
different, 7.4 wants FIXED values in ", whereas 6.2 does not enclose them in " like 
char fields).

Do you have any quick solution for me ?

Thanks,
Marco

_______________________________________________
sapdb.general mailing list
[EMAIL PROTECTED]
http://listserv.sap.com/mailman/listinfo/sapdb.general

Reply via email to