I can do this, which works. But the redundancy bothers me, and is prone to finger-check errors.
with pre_process as ( select recid, z_num, zip, zip4, dpbc, case when piecerate in ('AF','RF') and version_id = '81' then '81' else segment end as segment, ... blah blah ... from address_txt ) select printf( '%-1s%-15s%-5s%-4s%-2s%-1s%-4s%-4s%-1s%-5s%-1s%-10s%-10s%-10s%-1s%-20s%-20s%-15s%-1s%-10s%-20s%-20s%-50s%-50s%-50s%-50s%-50s%-25s%-2s%-8s%-003s%-006s%-009s', recid, z_num, zip, zip4, dpbc, segment ... blah blah ... ) from pre_process limit 10 ; On Thu, Oct 13, 2016 at 2:02 PM, Don V Nielsen <donvniel...@gmail.com> wrote: > These are simply blanks, 0x20, use to create separation of the output > columns. I'm assuming this is an inherent behavior for readability. If the > output was not being directed to the output file, it would be directed to > the display. > > I'm trying to avoid pre processing (creating a table or view of the > preprocessed data) and post processing (having to pass 10g of text data to > removed blanks -- which is dangerous on its own.) > > dvn > > On Thu, Oct 13, 2016 at 1:56 PM, Simon Slavin <slav...@bigfraud.org> > wrote: > >> >> On 13 Oct 2016, at 7:51pm, Don V Nielsen <donvniel...@gmail.com> wrote: >> >> > Unfortunately, there are two blanks separating each column >> >> Can you tell what characters these are ? Perhaps use a hexdump facility. >> >> My guess at this point is that you should continue with the file you have >> already developed and then post-process it to remove the blank characters. >> If those characters are used only in the blanks you don't want, it should >> be possible to use a simple find/replace utility to do it. >> >> Simon. >> _______________________________________________ >> sqlite-users mailing list >> firstname.lastname@example.org >> http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users >> > > _______________________________________________ sqlite-users mailing list email@example.com http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users