Shell Script to Delete line if pattern exists

ghostdog74 · 12-26-2007, 09:29 AM

you should start a new thread.

mvona · 09-02-2010, 10:24 AM

Quote:

Originally Posted by Anant Khaitan

Guys please tell me how do I delete a particular string.. Like for

Code:

echo abcdef.abc | <additional command>

should return only

Code:

abcdef

i.e delete everything after the dot..
Is it possible with tr

This will replace all occurrences of '.abc' with null.

Code:

echo abcdef.abc | sed 's/.abc//g'

admod · 09-22-2010, 01:07 AM

Hi pal,
for do my work, where I need to work all files in a folder(which are .js or .html or .php extension only), I work a script like this,
here my pattern is "123". Which delete a line which encounter a "123"
any where in line , the script is:

#!/bin/bash
grep -rl --include=*.{php,js,html} 123 $1 | xargs sed -i".bckp" '/123/d'
rm -f $1/*.bckp
#echo $1
#echo $2

Wish this may help you
Regards
Anes P.A

kontrabant · 10-15-2010, 05:33 AM

1. option

echo abcdef.abc | cut -d"." -f1

2.option

basename abcdef.abc .abc

3.option

x=abcdef.abc
echo ${x%.*}
echo ${x/.*/}

4.option

x=abcdef.abc
echo $x | perl -pe 's/\..*$//;s{^.*/}{}'

So there you have something to work with.

Code:

:-)=

Lucifer999 · 06-25-2011, 11:21 AM

Hi,
I am having similar problem.
I need to replace
username1 with #username1
username2 with #username2
and so on
in 800 different text files in different directories.

markush · 06-25-2011, 02:09 PM

Hello Lucifer999, welcome to LQ,

please open your own thread for this question. This one is quite old and most of the posters are no longer active here in the forums.

Markus

Lucifer999 · 06-26-2011, 01:42 AM

Thank you for the suggestion.....

cristalp · 08-23-2011, 04:58 AM

Quote:

Originally Posted by jkobrien

awk 'match($0,"username@email.com") == 0 {print $0}' FILENAME > NEWFILENAME

John

Here is another version:

Code:

awk '$0 !~ /username@email.com/ {print}' FILENAME > NEWFILENAME

In fact it is the same as yours. I found all awk programming share same logic, isn't it? But awk always looks not simple enough, a bit heavy. Can it be easier as Sed?