Deleting one pattern from a file

[rawatds]

Hi all
i want a unix script which will accept a string from user and then find that pattern in all files present in that
path and if pattern finds then remove that pattern the file.

e.g Let us suppose the lines in the files are
i am going to US
i am going to UK
UK is beautiful

so if my pattern is UK
so the new file should be :
i am going to US
i am going to UK
is beautiful

Thanx in advance
rawat

 

[PHV]

man find
man sed

Why is 'i am going to UK' unchanged ?

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ222-2244

 

[rawatds]

Hi
Can I use following code :
grep $string filename
and then can use sed 's/$string//g' . Please advice

rawat

 

[rawatds]

Hi all
i wrote this script :

echo " enter the string to delete "
read pattern
for i in `ls`
do
grep $pattern $i
if [ $? -eq 0 ]
then
sed 's/$pattern//g' $i > $i.1
fi

But it didnot work
till sed the loop is going but i donot know if we can
pass $pattern in sed .
any help ???

regards
rawat

 

[PHV]

you may try this:
echo " enter the string to delete "
read pattern
for i in `ls`
do
grep -q "$pattern" $i && sed "s/$pattern//g" $i > $i.1
done

In your shell man page take a look at the quoting mechanism.

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ222-2244