lat@stcvax.UUCP (Larry Tepper) (10/11/84)
I'm tempted to post this to net.bugs.4bsd... Call me naive if you wish, but why is /usr/ucb/head a program and not a shell script? It was my understanding that with UNIX you were supposed to build new tools based upon current ones, and that the shell was also considered a programming language. In addition, I foolishly thought that if a function is not used FREQUENTLY, then the "overhead" of using a shell script was considered acceptable. Perhaps there are people out there who are head heads??? (Ouch! 8-) I believe the following shell script is equivalent to /usr/ucb/head. It could even be 5 lines shorter if, when presented with several files, you were willing to accept an extra newline at the start of the first one. : head -- give the first few lines of a stream count=10 case $1 in -[0-9]*) count=`echo $1 | sed s/-//` shift ;; esac case $# in 0|1) sed ${count}q $1 exit $? ;; esac : first file is special sed "1i\\ ==> $1 <== ${count}q" $1 shift for f in $* do sed "1i\\ \\ ==> $f <== ${count}q" $f done -- Violence is the last refuge of the incompetent. {ihnp4 hao philabs sdcrdcf ucbvax!nbires}!stcvax!lat Larry Tepper Storage Technology, MD-3T, Louisville, CO 80028 303-673-5435
rik@uf-csg.UUCP (Rik Faith [guest]) (10/29/84)
<> Certainly many current standard UN*X tools can be written as shell scripts. The disk memory overhead required to maintain a binary, however, seems minimal, especially on a system with lots of disk space, when compared with the time and process overhead of a shell script. The tools, as binaries, are truly shell independent, as I feel they should be: as a user of csh, I am constantly spawning sh processes to run shell-scripts that could have been written in C. -- Rik Faith, student at the University of Florida, Gainesville UUCP: ..!akgua!uf-csv!uf-csg!rik [Is not life a hundred times to short for us to bore ourselves? -F.N.]