Output only MAC address on Ubuntu

You can access the address file for each device on the /sys virtual filesystem. The MAC address should be in /sys/class/net/<device-name>/address:

$ cat /sys/class/net/enp1s0/address
34:17:eb:5d:88:7c

For all devices:

$ cat /sys/class/net/*/address
34:17:eb:5d:88:7c
00:00:00:00:00:00
64:5a:04:69:50:45

The easiest way would be to use grep with PCRE:

$ ifconfig -a | grep -Po 'HWaddr \K.*$'
74:d4:35:84:34:13  

• grep -P will enable us to use perl compatible Regex

• grep -o will only take the matched portion of the line

• We have matched HWaddr before our desired match (MAC addresses) and then discard HWaddr by \K to print only the MAC addresses.

@Helio has mentioned an important point, this is highly dependent on your language i.e. locale settings. To overcome that you can use the C locale (uses ASCII character set) for this command only:

$ LANG=C ifconfig -a | grep -Po 'HWaddr \K.*$'
74:d4:35:84:34:13  

Here are a few ways:

1. grep. There are various regular expressions that will pick these up. Here, I am looking for 5 repetitions of 2 letters or numbers followed by a colon, then any two characters. The -i makes the match case insensitive and the -o makes grep print only the matched portion. -E enables extended regular expressions. The same regex also works with PCREs (-P).

   ifconfig -a | grep -ioE '([a-z0-9]{2}:){5}..'
   

2. sed. The -n suppresses normal output and the -r enables extended regular expressions. Using the same regex as above, this script will attempt to replace everything on the line with the part of it that matches the regex. If the substitution was successful, the resulting line is printed (because of the p at the end of the substitution).

   ifconfig -a | sed -rn 's/.*(([a-z0-9]{2}:){5}..).*/\1/p'
   

3. awk. If the line starts with a word character ([a-zA-Z0-9_]), and has 5 fields, print the last one.

   ifconfig -a | awk '/^\w/&&NF==5{print $NF}'
   

4. Perl, where, as usual, there are more than one ways to do it. This one is the same logic as the awk above. The -a tells perl to split each input line into the @F array.

   ifconfig -a | perl -lane 'if(/^\w/&&$#F==4){print $F[$#F]}'
   

   Alternatively, you can use the regex from the previous approaches:

   ifconfig -a | perl -lne '/(([a-z0-9]{2}:){5}..)/ && print $1'
   

5. Coreutils.

   LANG_ALL=C ifconfig -a | grep 'HWadd' | tr -s ' ' '\t' | cut -f 5