I had a regex that correctly matched a subset of data I was processing. When I ran it against the full data set, it started failing. I noticed some values were 'N A' versus either a MAC address or AP name, depending upon the column.
Sample data:
00:0b:85:57:bc:c0 00:0b:85:57:bc:c1 AP1130 10.10.163.217 Joined
00:1c:0f:81:db:80 00:1c:63:23:ac:a0 AP1140 10.10.163.216 Joined
00:1c:0f:81:fc:20 00:1b:d5:9f:7d:b2 AP1 10.10.163.215 Joined
00:1c:0f:81:fc:20 N A N A 10.10.163.215 Not joined
00:21:1b:ea:36:60 00:0c:d4:8a:6b:c1 AP2 10.10.163.214 Joined
Regexp:
((?:(?:[0-9a-f]{2}[:-]){5})(?:[0-9a-f]{2}))(?:\s+?)(((?:(?:[0-9a-f]{2}[:-]){5})(?:[0-9a-f]{2}))|(N A))(?:\s+)((AP.+?)|(N A))(?:\s)
I have modified my regex but it still isn't matching the MAC address or 'N A'. Same for the name field: Match the AP name or 'N A'
My work as it stands: https://regex101.com/r/sgGEzh/1
I assume my brackets are not correct, but I can't see where my (|) OR operator is failing. I am getting duplication of some groups now.
I should match the first MAC address, the second MAC address or the string 'N A', and last the AP name or the string 'N A'.
I should always get three matching groups per line.
I was until I tried to process the 'N A' strings.
You're matching the correct things, you just need to get rid of the unneeded capturing groups around N A and AP.+?. These are causing these strings to end up in different groups in the result. You only need 3 capturing groups.
You have a number of other groups that aren't really needed, like the non-capturing group around \s+?. You don't need a group around each | alternative if they're already inside a group. The only non-capturing group you need is the one around [0-9a-f]{2}[:-] when it's being quantified.
The following works and removes all the redundant groups:
((?:[0-9a-f]{2}[:-]){5}[0-9a-f]{2})\s+?((?:[0-9a-f]{2}[:-]){5}[0-9a-f]{2}|N A)\s+(AP.+?|N A)\s