Remove last part of a text line in a file

[bcooperizcool]

How can I remove the last part of lines from a file?

my code that doesn't work



while read name; do
sed "s/$name.app$//" < fileone | sort | uniq > filetwo
done < filewiththingtoremove


As always, Thanks!

[slavik]

you also need to add .* to the part you want to remove, unless you don't want to wildcard stuff, can you give us a short example of text and what you are trying to do?

[bcooperizcool]

where do I need to add .*?

Example of file


/var/stash/Applications/245-234453-GFSD/Facebook.app
/var/stash/Applications/PSL-742343-NWIM/Text.app


Example of file that contains the stuff to remove


Facebook
Text


Example of hopeful output into new file


/var/stash/Applications/245-234453-GFSD
/var/stash/Applications/PSL-742343-NWIM

[worksofcraft]

Well if you don't mind compiling it...



// g++ pathsplit.cpp
#include <cstdio>

char *split_path(char *buf) {
char *p, *name;
p = name = buf;
// scan command line for /
while (*p) {
if (*p == '/') name = p; // leave a marker on the last /
++p;
}
if (name == buf) return NULL;
*name++ = '\0'; // string terminator
return name;
}

int main() {
char buf[2020];
while (fgets(buf, sizeof(buf), stdin)) {
if (split_path(buf)) printf("%s\n", buf);
}
return 0;
}

[Vaphell]

echo '/var/stash/Applications/245-234453-GFSD/Facebook.app' | sed -r 's:/[^/]+app$::'
output

/var/stash/Applications/245-234453-GFSD

'/', one-or-more non-slash chars, 'app', end of line in case you wonder, could be even reduced to /[^/]*$

[bcooperizcool]

Ehhhh.... its for an iphone. I don't think it supports that. :( I found another way by recoding some of the big script Thanks though!

[kurum!]

ruby -F'/' -ane 'BEGIN{f2=open("file2").read.split("\n")}; puts $F[0..-2].join("/") if f2.grep($F[-1].split(".")[0]).size>0 ' file

[Arndt]

Well if you don't mind compiling it...



// g++ pathsplit.cpp
#include <cstdio>

char *split_path(char *buf) {
char *p, *name;
p = name = buf;
// scan command line for /
while (*p) {
if (*p == '/') name = p; // leave a marker on the last /
++p;
}
if (name == buf) return NULL;
*name++ = '\0'; // string terminator
return name;
}

int main() {
char buf[2020];
while (fgets(buf, sizeof(buf), stdin)) {
if (split_path(buf)) printf("%s\n", buf);
}
return 0;
}


You could probably use 'strrchr' somewhere here.

[geirha]

awk -F/ 'BEGIN{OFS=FS} NR==FNR{a[$0".app"];next} $NF in a {NF--} 1' filewiththingtoremove fileone > filetwo