Re: Qshell and CCSID

Hi Kelly,

> My questions are:
> (1) Can I set an environmental variable or something
> so Qshell defaults to ASCII instead of EBCDIC when it
> creates stream files?

Each stream file is tagged with it's own CCSID.  The Qshell utilities (and
in fact, most OS/400 utilities) will automatically translate from & to
that CCSID.  (Naturally, the CCSID defaults to that of your job, which for
most everyone is EBCDIC.)

The "touch" utility in QShell can be used to create an empty file and
assign it a CCSID.  For example:

   rm outfile.txt                   (get rid of old copy w/wrong CCSID)
   touch -C 1252 outfile.txt        (create w/windows 1252 CCSID.
                                     The -C MUST be captialized)
   grep "whatever" infile.txt > outfile.txt

At this point, outfile.txt will be in CCSID 1252 (which is the most common
ASCII used on Windows)


> (2) Can I configure something so Qshell puts a CRLF at
> the end of lines instead of line feeds or tabs?

Because QShell is a unix-like interface, it uses the unix convention for
ends of lines, which is "line feed" only (though unix people usually call
them 'newlines')

Off the top of my head, I can't think of an easy way to write out the
extra CR. I can think of some complicated ones, but let's try something
else first:

One very easy solution to this is to open the file in Wordpad instead of
Notepad :)  Wordpad understands Unix end-of-line sequences and handles
them nicely.


> (3) Is there a Qshell utility that would let me find
> and replace character sequences in stream files such
> that I could replace all line feeds (or tabs) with
> CRLF?

Uhhh...  There's sed, but the syntax is rather ugly.  You could do this:

   rm outfile.txt
   touch -C 1252 outfile.txt
   grep "." infile.txt | sed "s/$/$(printf \\\\\\r)/g"> outfile.txt

I imagine there's an easier way, but I can't think of what it might be...