ansaurus

Question

Answer 1

A:

You can rewrite the file with some editor(like nano, vi, vim, etc.) or you can follow the instructions of this website:

http://www.liamdelahunty.com/tips/linux_search_and_replace_multiple_files.php

Or replacing a string with a simple grep, like this:

http://www.debianadmin.com/howto-replace-multiple-file-text-string-in-linux.html

When I say rewrite, I mean replace ":-" for ""...

Fredy87 2010-09-21 13:24:49

Answer 2

+4 A:

sed 's#db/fs-type$##' myfile > myalteredfile

sleepynate 2010-09-21 13:25:33

+1, beat me to it by a few seconds

ammoQ 2010-09-21 13:26:38

please you can give me example

Osama Ahmad 2010-09-21 13:26:50

your sed does not remove all occurences of the string in every line of the file, add the 'g' switch...

eumiro 2010-09-21 13:31:29

@eumiro It is obvious from the examples he wants to remove them from the end of the line.

sleepynate 2010-09-21 14:17:39

@Mohammed this is an example. If the information were in `myfile`, you'd use the `sed` command as above in the same directory, and it would generate `myalteredfile`, a copy with your changes.

sleepynate 2010-09-21 14:18:39

and we can use cat FILENAME | sed -e 's/db\/fs-type//g'

Osama Ahmad 2010-09-21 14:18:43

@sleepynate: It seems obvious from the example that the OP wants to remove "db/fs-type" from *everywhere* in the line (there are three in the original and none in the result). Furthermore, your answer removes the text, but the example shows spaces replacing the db/fs-type.

GreenMatt 2010-09-21 14:24:05

@GreenMatt: I think you're looking at the previous malformed version of the question. There are three lines in the example, each only containing one instance of the text to be removed, occurring at the end of the line.

sje397 2010-09-21 23:32:23

@GreenMatt yea, I think you may be looking at a messed up version of the example. I don't see what you're describing in the current example.

sleepynate 2010-09-22 03:39:40

@Mohammad AL-Rawabdeh, @sje397, @sleepynate: I've looked at the original and both edits and copied it would be nice if Mohammad could clarify what is wanted, instead of making us guess.

GreenMatt 2010-09-22 14:14:02

@GreenMatt well mate, we're all here to do our best and help each other out. *group high-five*

sleepynate 2010-09-22 18:12:59

Answer 3

A:

I would probably use sed. After some experimentation I got the following

sed 's:/db/fs-type:/ :g' path.txt

to produce

/var/lib/svn/repos/b1me/products/payone/generic/code/core/ /var/lib/svn/repos/b1me/products/payone/generic/code/fees/ /var/lib/svn/repos/b1me/

which seems to be the output you want, assuming path.txt contains the original path you wish to make changes to.

If you want to capture this in another file, the command would be something like:

sed 's:/db/fs-type:/ :g' path.txt > new_path.txt

GreenMatt 2010-09-21 13:27:48

Answer 4

A:

so ross$ a=/x/y/db/fs-type
so ross$ b=${a%/db/fs-type}
so ross$ echo $b
/x/y

So now to do the whole file...

so ross$ expand < dbf.sh
#!/bin/sh
t=/db/fs-type
while read f1 f2 f3; do
  echo ${f1%$t} ${f2%$t} ${f3%$t}
done
so ross$ cat dbf
/a/b/db/fs-type /c/d/db/fs-type /e/f/db/fs-type
/a/b/db/fs-type /c/d/db/fs-type /e/f/db/fs-type
/a/b/db/fs-type /c/d/db/fs-type /e/f/db/fs-type
/a/b/db/fs-type /c/d/db/fs-type /e/f/db/fs-type
so ross$ sh dbf.sh < dbf
/a/b /c/d /e/f
/a/b /c/d /e/f
/a/b /c/d /e/f
/a/b /c/d /e/f
so ross$

DigitalRoss 2010-09-21 13:32:25

Answer 5

A:

One in perl :)

cat oldfile.txt | perl -nle "s/db\/fs-type//g; print;" > newfile.txt

Powertieke 2010-09-21 14:07:38

ansaurus

tags:

views:

answers:

how i can delete something from file ?

related questions