UNIX for Dummies Questions & Answers

Hello,

i want built a log analyzer for nginx. Okay and i use it as training for the shell tools. The most what i want i could relize. But i has trouble with dynamic things. I have the IP address extracted and has set the geo localtion for the ip.

I would like to count the countries. With fixed data, no problem. But how can I automate it without giving a pattern. For example i often has china, germany and united states. But im to has some brazil, russia or algeria.

So i grep with pattern:

Code:

tr -s ' ' '\n' < iplog.txt | grep -c China

But i want grep and count automatic when is more then zero entry in file for a country?

The file is built ip country.

Has someone an idea? I hope i has correct explain.

Thank you for help & Nice Day
Silvio

This could easily be done with the associative array feature of awk. What is the format of the file?

Suppose the country name is in column 2 then you could use something like:

Code:

awk '{A[$2]++} END{for(i in A) print i, A[i]}' iplog.txt

Hello,

the file is so:

Code:

ipadress	 Russian Federation
ipadress	 Germany
ipadress	 United States
ipadress	 United States
ipadress	 Brazil
ipadress	 China

I think i can use:

Code:

cat ip.txt | tr -cs A-Za-z\' '\n'| tr A-Z a-z | sort | uniq -c > ipstatstemp.txt

So it work.

THank you for help & Nice Day

Silvio

This should work then:

Code:

awk '{A[$2]++} END{for(i in A) print A[i], i}' FS='\t' OFS='\t' iplog.txt