CODEWITHSUNDEEP
bashawkgrep
Areza
Areza

Reputation: 6080

counting number a certain word per line in a file

I have a file looks like

Old data

alireza,mahdi,alireza,mohammad
alireza,olivia,amireza,hasan
alireza,alireza,alireza,alireza

New data

0.54:30.36:N    0.54:32.31:N     0.54:30.36:Y
0.54:30.36:Y    0.54:32.31:N     0.54:30.36:Y
0.54:30.36:N    0.54:32.31:N     0.54:30.36:Y
0.54:30.36:N    0.54:32.31:N     0.54:30.36:N

and I would like to count number of say, "Y" per line,

so the output would be 

1
2
1
0

so I can grep it like `"grep -w ':Y' myfile" but then I don't know how to count it ! since wc ...

**UPDATE I am sorry to change the data and possibly the pattern. I wanted to present a simple and understandable format which I was not successful. Again, I do apologize.

Upvotes: 0

Views: 101

Answers (2)

Ed Morton
Ed Morton

Reputation: 204628

This is the idiomatic awk way to print how may times a RE occurs on each line:

$ awk '{print gsub(/:Y/,"")}' file
1
2
1
0

Upvotes: 6

devnull
devnull

Reputation: 123658

For your new data, you could simply change the field separator:

awk -F'[ :.]' '{for(i=1;i<=NF;i++) {if ($i=="Y") c++;}}{print c; c=0}' inputfile

or make use of a pattern to match against:

awk '{for(i=1;i<=NF;i++) {if ($i~/:Y$/) c++;}}{print c; c=0}' inputfile

Both would yield:

1
2
1
0

for your new input.

Old answer: You could use awk:

awk -F, '{for(i=1;i<=NF;i++) {if ($i=="alireza") c++;}}{print c; c=0}' inputfile

For your sample input, it'd produce:

2
1
4
0

Upvotes: 0

Related Questions