markk@censor.UUCP (Mark Keating) (06/16/89)
After trying to work on another system I was slightly dissapointed
to find that there was no 'path' program to be found anywhere.
So I wrote one -- hope you find it useful.
#! /bin/sh
# This is a shell archive, meaning:
# 1. Remove everything above the #! /bin/sh line.
# 2. Save the resulting text in a file.
# 3. Execute the file with /bin/sh (not csh) to create the files:
# path.1
# path.c
# This archive created: Thu Jun 15 16:04:06 1989
export PATH; PATH=/bin:$PATH
if test -f 'path.1'
then
echo shar: will not over-write existing file "'path.1'"
else
cat << \SHAR_EOF > 'path.1'
.TH PATH 1
.SH NAME
path \- locate the executable for a command
.SH SYNOPSIS
/usr/local/bin/path [ -options ] command(s)
.I file
.SH DESCRIPTION
.I Path
will report on the location of
.I file,
searching every directory in the path (as defined in the
$PATH variable), much like the operating system does when searching
$PATH to execute a program. It is handy for finding out where in your path
a command resides. Any options specified are passed with located names to the
.I ls
command.
.PP
.SH CREDITS
This utility was inspired by the
.I findpath
utility posted by Ed Carp,
and re-generated into
.I path
by Mark Keating.
SHAR_EOF
fi # end of overwriting check
if test -f 'path.c'
then
echo shar: will not over-write existing file "'path.c'"
else
cat << \SHAR_EOF > 'path.c'
/* opt: -O -s -o path
path [ -options ] command [ command... ]
find the path of a command
if options specified then they are passed through to `ls'
path Version 1.0 created by Mark Keating - 1989
*/
#include <stdio.h>
#include <string.h>
#include <sys/types.h>
#include <dirent.h>
DIR *d_fp;
struct dirent *d_ent;
char *d_nam;
char *ls_opt, ls_buf[256];
char *path, path_list[512];
char *getenv();
main(argc, argv)
int argc;
char *argv[];
{
register int i, found;
register char *ptr;
if (argc < 2) {
fprintf(stderr, "Usage: path [-lsopt] command [...]\n");
exit(1);
}
if ((path=getenv("PATH")) == NULL) {
fprintf(stderr, "PATH not set!\n");
exit(1);
}
for (i=1; i < argc; i++) {
if (argv[i][0] == '-') { /* ls option passed */
ls_opt = argv[i]; continue;
}
strcpy(path_list, ".:"); strcat(path_list, path);
for (ptr = path_list, found = 0;
!found && (d_nam=strtok(ptr, ":")) != NULL; ptr = NULL) {
if ((d_fp=opendir(d_nam)) == (DIR *)NULL) {
perror(d_nam); continue;
}
while ((d_ent=readdir(d_fp)) != (struct dirent *)NULL) {
if (strcmp(d_ent->d_name, argv[i]) == 0) {
if (ls_opt == NULL) {
printf("%s/%s\n", d_nam, argv[i]);
}
else {
sprintf(ls_buf, "ls %s %s/%s\n",
ls_opt, d_nam, argv[i]);
system(ls_buf);
}
found = 1; break;
}
}
closedir(d_fp);
}
}
}
SHAR_EOF
fi # end of overwriting check
# End of shell archive
exit 0mpl@cbnewsl.ATT.COM (michael.p.lindner) (06/16/89)
In article <702@censor.UUCP>, markk@censor.UUCP (Mark Keating) writes: > After trying to work on another system I was slightly dissapointed > to find that there was no 'path' program to be found anywhere. > So I wrote one -- hope you find it useful. I have a program which does just about the same thing. It is written in shell. The difference is it takes an optional path or paths, so it can also be used for finding things like "where is this #include file found" or "where is this source file found (by using MAKEPATH). Here it is. Enjoy! Mike Lindner attunix!mpl #! /bin/sh # This is a shell archive, meaning: # 1. Remove everything above the #! /bin/sh line. # 2. Save the resulting text in a file. # 3. Execute the file with /bin/sh (not csh) to create the files: # whf.1 # whf.sh # This archive created: Fri Jun 16 09:04:06 1989 export PATH; PATH=/bin:$PATH if test -f 'whf.1' then echo shar: will not over-write existing file "'whf.1'" else cat << \SHAR_EOF > 'whf.1' .TH WHF 1 .SH NAME whf \- locate a file using a path .SH SYNOPSIS whf file [path ...] .SH DESCRIPTION .I Whf will print the location of file by searching each directory in the given path(s). If no path is specified, $PATH is used. If one or more paths are given, each should be a list of colon separated directories, and are searched left-to-right. .PP .SH CREDITS This utility is functionally an extension of the .I whf command I had on my home machine at AT&T Bell Labs. The source is my own invention, and is not based on anything. SHAR_EOF fi # end of overwriting check if test -f 'whf.sh' then echo shar: will not over-write existing file "'whf.sh'" else cat << \SHAR_EOF > 'whf.sh' # whf - find a file somewhere # by Michael Lindner IFS=":$IFS" case "$#" in 0) echo >&2 "usage: $0 file [path ... ]" exit 1 ;; 1) file="$1" shift for d in $PATH do if [ -s $d/$file ] then echo $d/$file exit 0 fi done ;; *) file="$1" shift for d in $* do if [ -s $d/$file ] then echo $d/$file exit 0 fi done ;; esac echo >&2 "$0: $file not found" exit 1 SHAR_EOF fi # end of overwriting check # End of shell archive exit 0
rustcat@csli.Stanford.EDU (Vallury Prabhakar) (06/17/89)
In article <702@censor.UUCP> markk@censor.UUCP (Mark Keating) writes:
%
% path \- locate the executable for a command
[...]
% will report on the location of
% .I file,
% searching every directory in the path (as defined in the
% $PATH variable), much like the operating system does when searching
% $PATH to execute a program. It is handy for finding out where in your path
% a command resides.
[...Source code deleted...]
Is this in any way better or different from the `which' command on BSD
systems? Just curious. markk@censor.UUCP (Mark Keating) (06/19/89)
I've received several cards 'n letters from some of the folks out there in netland, about my recent path posting, and no I don't know how it differs, if at all to the BSC which program, any clarification here would be appreciated, although I don't think it's a SYSV thing, I know we don't have it (which) here. I also recieved a script, and one was posted as well to do the same thing with the exception that path entries located are not conditionally passed through `ls', which I find very useful ie: `path path -l path -x path' will produce /usr/bin/path -rwxr-xr-x 1 root root 11896 Jun 17 1989 /usr/bin/path /usr/bin/path all the options are passed through 'ls' so more extravagent combinations could be produced, most useful is 'path -l name', but you probably get the idea. And yes I know that "ls -l `path path`" would work as well, just not as convienently or allow changing options along the line. The thing I liked about the one script was to specify the path, BUT since everything that is an option, is blindly passed to ls for simplicity, anybody have a good suggestion on the best way to handle this ?? Hope I haven't started wars here, or put us on a which path. Appreciate the Response. Mark Keating - markk@censor
mpl@cbnewsl.ATT.COM (michael.p.lindner) (06/20/89)
In article <720@censor.UUCP>, markk@censor.UUCP (Mark Keating) writes: > > I also recieved a script, and one was posted as well to do the > same thing with the exception that path entries located are > not conditionally passed through `ls', which I find very useful > > The thing I liked about the one script was to specify the path, BUT since > everything that is an option, is blindly passed to ls for simplicity, > anybody have a good suggestion on the best way to handle this ?? simple! Here's the revised script: options= while [ "$#" -gt 0 ] do case "$1" in --) shift break ;; -*) options="$options $1" shift ;; *) break ;; esac done IFS=":$IFS" case "$#" in 0) echo >&2 "usage: $0 file [path ... ]" exit 1 ;; 1) file="$1" shift for d in $PATH do if [ -s $d/$file ] then /bin/ls$options $d/$file exit 0 fi done ;; *) file="$1" shift for d in $* do if [ -s $d/$file ] then /bin/ls$options $d/$file exit 0 fi done ;; esac echo >&2 "$0: $file not found" exit 1