CODEWITHSUNDEEP
regexawksedfield
Ghost
Ghost

Reputation: 1523

How can I get the field number that contains a MAC address with sed or awk?

I'm trying to get the field number that contains a certain pattern (in this case a MAC address). Imagine we have something like this:

AAA BBB CCC xx:xx:xx:xx:xx:xx EEE FFF GGG

Where xx:xx... is a MAC address, so x's are alphanumeric.

If I use that line as the input, I want to obtain the number of the field where the MAC address is. In this case, "4".

I was thinking of sed or awk, but don´t know how to output the field number.

Upvotes: 1

Views: 236

Answers (2)

Jason Hu
Jason Hu

Reputation: 6333

sed would work:

echo YOUR_STRING | sed -n 's/.*\([a-zA-Z0-9]\{2\}\(:[a-zA-Z0-9]\{2\}\)\{5\}\).*/\1/p'

Example: 

echo abc ab:ab:ab:ab:ab:ab abc | sed -n 's/.*\([a-zA-Z0-9]\{2\}\(:[a-zA-Z0-9]\{2\}\)\{5\}\).*/\1/p'

ab:ab:ab:ab:ab:ab

Upvotes: 1

Todd A. Jacobs
Todd A. Jacobs

Reputation: 84393

Loop Over Fields With NF

Use a for-loop to check each field against a regular expression, using the POSIX character class for hexadecimal digits. When a match is found, print the matching field number. 

echo 'AAA BBB CCC 00:00:00:00:00:00 EEE FFF GGG' |
    awk '{ for (i=1; i<=NF; i++)
               if ($i ~ /^([[:xdigit:]]{2}:){5}[[:xdigit:]]$/)
                   print i
                   next
         }'
4

Upvotes: 2

Related Questions