On 5/13/08, Dr. Werner Fink <[EMAIL PROTECTED]> wrote:
> On Tue, May 13, 2008 at 01:18:40PM +0200, William James wrote:
> > Imagine I have a file like this:
> >
> > buf="$(cat <<EOF
> > part1: a part
> > part5: another part
> > # comment
> > part9: 9th part
> > EOF
> > )"
> >
> > How can I get the line containing 'part5' using ksh93 string
> > operators, avoiding both 'while read...' and 'sed'? I tried
> > ${buf##pattern}, ${buf#pattern}, ${buf%pattern}, ${buf%%pattern} but
> > they all _remove_ the string pattern. I need a string operator which
> > only leaves the string matching pattern and removes the rest.
>
>
> Not a string operator but this seems to work:
>
>
> buf="$(cat <<EOF
> part1: a part
> part5: another part
> # comment
> part9: 9th part
> EOF
> )"
>
>
> OIFS="$IFS"
> IFS=$'\n'
> lines=(${buf})
> IFS="$OIFS"
>
> typeset -i [EMAIL PROTECTED]
> while let i-- ; do
> case "${lines[$i]}" in
> part5:*)
> echo "${lines[$i]}"
> break
> ;;
> esac
> done
>
> on the other hand, a read loop within a pipe or simply a loop
> with reading from stdin on the file seems much faster:
>
> while read line ; do
> case "$line" in
> part5:*)
> echo "$line"
> break
> ;;
> esac
> done <<EOF
I wish to avoid using read. /bin/sed can cut and paste fields based on
pattern and I'm searching for something similar in ksh93, i.e. the
reverse operation of ${buf##pattern}, instead of removing pattern
remove all characters which do not match pattern.
Cheers,
William
--
@,,@ William James
(\--/) [EMAIL PROTECTED]
(.>__<.) GNU/Solaris hacker
_______________________________________________
ast-users mailing list
[email protected]
https://mailman.research.att.com/mailman/listinfo/ast-users
