Re: Unload command help

In article <6lpdfk$b1e1@webint.na.informix.com>, no_spam_please@we_are_british.com wrote: > > How about creating a file called, say, "do" which contains the following:- > > isql nsoe - << EOF > select field1 from table;> EOF > > and then saying the following at the unix prompt:- > > do | sed 's/$/ /' | cut -c1-20 > outputfile > > That is, never mind unload; just take the output of "select field1 from table" > and pipe it through "sed" to add twenty blanks to the end of each line, then > pipe that through "cut" to get the first twenty characters in each row. Seems > like it should work. > > You'll need to edit the first few lines out of "outputfile" because "do" > will output > > +------------- > | > |field1 > | > |000AAABB > |000AAAC > |000AAAD > |... > > rather than > > +------------- > |000AAABB > |000AAAC > |000AAAD > |... > > - Paul (not a spokesman) I did not try it, but let me add: when you use OUTPUT TO PIPE "mysedscript.sh" and mysedscript.sh is a script containing Pauls's sed command and when you also add WITHOUT HEADINGS you will not get the unwanted line containing the header "field1" , i.e., if the file called do contains isql nsoe - << EOF OUTPUT TO PIPE "mysedscript.sh" WITHOUT HEADINGS select field1 from table;EOF (Then you maybe still have one blank line at the beginning,but I did not test this.) It makes things more compact Dick > > article <35802FEF.C1ABA565@mci.com>, David Hinz <David.Hinz@mci.com> wrote: > >I need help with an unload command to do the following: > > > >- The table has one field: field1 char(10) > >- All the records in the table look like "0000000ABC" or "0000000AB" > >> [ ..... for shortness ..... ] -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading