CODEWITHSUNDEEP
awk
user2464553
user2464553

Reputation: 85

awk or sed to delete a pattern with string and number

I have a file with the following content:

string1_204
string2_408
string35_592

I need to get rid of string1_,string2_,string35_ and so on and add 204,408,592 to get a value. so the output should be 1204.

I can take out string1_ and string 2_ but for string35_592 I have 5_592. I cant seem to get the command right to do what I want to do. Please any help is appreciated:)

Upvotes: 1

Views: 975

Answers (2)

Thor
Thor

Reputation: 47109

In case you are interested in a coreutils/bc alternative:

<infile cut -d_ -f2 | paste -sd+ - | bc

Output:

1024

Explanation:

cut splits each line at underscore characters (-d_) and outputs only the second field (-f2). The column of numbers is passed on to paste which joins them on a line (-s) delimited by plus characters (-d+). This is passed on to bc which calculates and outputs the sum.

Upvotes: 3

hek2mgl
hek2mgl

Reputation: 158020

With awk:

awk -F_ '{s+=$2}END{print s}' your.txt

Output:

1204

Explanation:

-F_    sets the field separator to _ what makes it easy to access
       the numbers later on

{
    # runs on every line of the input file
    # adds the value of the second field - the number - to s.
    # awk auto initializes s with 0 on it's first usage
    s+=$2
}
END {
    # runs after all input has been processed
    # prints the sum
    print s
}

Upvotes: 5

Related Questions