in reply to unhead and untail

A couple of thoughts:

The cutting up of input stream as you describe can be much done much easier with sed, for instance, instead of

query-students | sort | unhead '^D' | untail '^G'
do
query-students | sort | sed '/^D/,/^G/p; d'
Also, neither head nor tail modify files they operate on. I understand, you were probably solving a specific problem, but this script can be generalized even more.

Replies are listed 'Best First'.
Re^2: unhead and untail (GNU sed -i)
by Aristotle (Chancellor) on Apr 15, 2003 at 20:09 UTC
    unhead is superfluous even with inplace editing given recent versions of GNU sed, as they sport an -i switch for inplace editing just like Perl has. What you still can't do with sed is specify "X lines before end of file" addresses.

    Makeshifts last the longest.

      > What you still can't do with sed is specify "X lines before end of file" addresses.

      Neither can this utility, read the description.