Script Help: How to count a matching pattern in one line?

dv502 · 12-13-2008, 11:28 AM

Hey everyone,

I know how to use grep, sed and awk in its most basic form, but I need to know how can I use either one of these commands or another to perform this task.

I Know how to count matching lines by piping it to wc. I want to know how can I count a matching pattern on one line, for example:

Quote:

The big bear jumped over the big fence.

I would like to know how many times the word big appears on one line, in our case 2 times.

In the past, I've used xargs to put each word on a separate line, then piped it to grep and wc to get the result.

If you know a better way I can achieve the same result please let me know.

Thanks in advance

colucix · 12-13-2008, 12:56 PM

Here is a method using awk. The gsub function returns the number of the substitution made, so you can simply do

Code:

echo "The big bear jumped over the big fence." | awk '{print gsub(/\ybig\y/,"")}'

The \y notation in the regular expression is the gawk-specific operator to match the word boundary. Use that if you want to exclude the pattern within other words and count only the word "big" as is.

Using the same method you can count the total number of a word in a file:

Code:

awk '{count += gsub(/\ybig\y/,"")} END{print count}' file

rizwanrafique · 12-13-2008, 01:02 PM

Code:

#!/bin/bash

COUNTER=0
LINE="The big bear jumped over the big fence"
for WORD in $LINE
do
    if [ "$WORD" = "big" ]
    then
        COUNTER=`expr $COUNTER + 1`
    fi
done

echo $COUNTER

dv502 · 12-13-2008, 01:53 PM

Thanks colucix, I tried both of your suggested codes and they work great and thank you rizwanrafique for the mini script you provided. The codes provided by both of you has help me greatly.

Thanks again.