Re: Finding members

[Vern Hamberg <vhamberg@xxxxxxxxxxxxxxxxxxxxxxxxx>]

Now who was complaining about black boxes? Joe, are you there?

Don't know about fixed-column formatiing, Syd. I wondered that, too. I
don't see anything in the doc that allows for formatting, but, as I say,
whole books are written on this. And the things that can be done are
related to how regular expressions work, I think.

Of course, you could create IFS output with certain delimiters, then use
CPYFRMIMPF to a physical file.

Ooh, this is neat stuff!

There is always awk, another command (or gawk, the GNU equivalent). I think
this is more advanced. It's available free from IBM in something called GNU
Utilities.

I DID find that I don't need to use so many pipes - there's a "-e" option
that appends commands together, so this works just like what I presented
before:

find /QSYS.LIB/VERN.LIB -name 'TEST*.MBR' | sed -e s\%/QSYS.LIB/%% -e
s/.LIB// -e s\%.FILE/%\(% -e /.MBR/s//\)/

Probably faster

Cheers

Vern

At 05:35 PM 11/8/02 +0000, you wrote:
> Thanks Vern,
>
> I can now see what is happening - Magic.
>
> I have been playing about with this and discovered it is possible to
> pipe direct to the QSYS.LIB file system:
>
> find Directory -name searchCriteria >
> /QSYS.LIB/OutputLib.LIB/OutputFile.FILE/OutputMbr.MBR
>
> If it is possible to output each part of the path as a fixed length,
> (eg. Each library, file or member is 10 chars. long), then it should be
> possible to create an outfile and associated logicals that can readily
> be used to display, print, etc.
>
> Alternatively %SCAN and %SUBST can be used on the data, but this
> restricts the possible uses of the output file.
>
> Is it possible to pad the library/file/member names with blanks to a
> fixed length?
>
> Syd