I have an output file where the string "mac address =" will always be on each line, but the location of the string may be at a different byte. For example:

SERVER-A <00> - IP ADDRESS = 192.168.10.10 <ACTIVE> mac address = 00-1A-A0-A0-D9-1D
mac address = EC-9A-74-64-E5-5D PRINTER-A.DOMAIN.COM - P <ACTIVE>

I want to strip the "-" out of the MAC Address only as there can be other "-" in each line. So I'd like the output to look like this.

SERVER-A <00> - IP ADDRESS = 192.168.10.10 <ACTIVE> mac address = 001AA0A0D91D
mac address = EC9A7464E55D PRINTER-A.DOMAIN.COM - P <ACTIVE>

I'm guessing that sed should be used, but have no idea how to achive the desired result.

Thanks in advance.

Define the data in the input, and what you want as output:
two alphanumerics, minus sign,two alphanumerics, minus sign, ...
two alphanumerics,two alphanumerics, ...

Sed supports character classes (with repeat counts) and back references to handle the above. What is your sed (regex really) comfort level ?.

Correct. I want the MAC Address minus the hyphen symbol but I also want to retain all the other information in each line. My comfort/knowledge of sed is a 3 on a scale of 10. I could do a simple sed 's/-//g' but that would also remove hyphens that are not a part of the MAC Address. Thank you.

There's probably a less messy solution, butyields the result you showed. I didn't actually match the "mac address =" string, just anything that had the pattern of 6 pairs of hex digits separated by minus signs.

1 members found this post helpful.

That worked! Thank you. I never would have figured that out.