Shell Programming and Scripting

Hi,

I have 10 different strings. I have to delete the whole line matching with any one string.

I can use sed like below

sed '/$keyword1/d' fileList.txt > temp_fileList.txt
sed '/$keyword2/d' temp_fileList.txt > temp_fileList1.txt
.
.
.

Here is the problem i do not have fixed number of strings and I want to eliminate the temp files creation.

Is there any way to achieve this.

Thanks,
Shekhar

If you have the strings in a file you can use grep with the -f and -v option.

Regards

Hi Franklin,

Can you tell me the syntax to use grep with -v & -f. I have strings to be matched in a file.

Code:

grep -v -f file_with_strings_in_it  fileList.txt

I tried with the below command

grep -v -f str.txt list.txt

Its not working, saying grep usage not allowed with -f.


Is there any other way to acheive the task.


Thanks,
Shekhar

Supposing the lines of list.txt match exactly the lines you want to delete in file.txt:

Code:

awk 'NR==FNR{a[$0];next} !($0 in a)' list.txt file.txt

Regards

in case your sed supports the in-line edit option:

Code:

for keyword in `cat str.txt`; do sed -i -e '/'$keyword'/d' list.txt; done