Regular expression assistance

agentorange has asked for the wisdom of the Perl Monks concerning the following question:

Hi
Struggling a bit with regular expressions and hopefully someone can help.
Parsing some Cisco CDP output and my string equals below:

$cdp = "Device-ID (0x01), length: 44 bytes: 'my-switch-name'
Address (0x02), length: 13 bytes: IPv4 (1) 10.1.1.10
Port-ID (0x03), length: 16 bytes: 'Ethernet101/1/20'
Capability (0x04), length: 4 bytes: (0x00000228): L2 Switch, IGMP snoo
+ping
Version String (0x05), length: 66 bytes:
  Cisco Nexus Operating System (NX-OS) Software, Version 5.2(1)N1(3)
Platform (0x06), length: 11 bytes: 'N5K-C5596UP'
Native VLAN ID (0x0a), length: 2 bytes: 911
AVVID trust bitmap (0x12), length: 1 byte: 0x00
AVVID untrusted ports CoS (0x13), length: 1 byte: 0x00
Duplex (0x0b), length: 1 byte: full
MTU (0x11), length: 4 bytes: 1500 bytes
System Name (0x14), length: 11 bytes: 'my-switch-name'
System Object ID (not decoded) (0x15), length: 14 bytes:
  0x0000:  060c 2b06 0104 0109 0c03 0103 880e
Management Addresses (0x16), length: 13 bytes: IPv4 (1) 10.151.69.11
Physical Location (0x17), length: 13 bytes: 0x00/snmplocation"
[download]

If using ksh and awk something like this would work:

VLAN=`echo $cdp | awk '/0x0a/ { print $8 }'`
[download]

And set VLAN to 911. But I just cannot see how to do it with Perl even though it is almost certainly ridiculously straight forward. my @switch = ( $cdp =~ m/0x01/ ); Isn't working and any assistance gratefully received.

Comment on Regular expression assistance Select or Download Code

Replies are listed 'Best First'.
Re: Regular expression assistance by kennethk (Abbot) on Dec 07, 2012 at 15:48 UTC
For me, the code `print "yes\n" if $cdp =~ /0x01/;` prints `yes` which means that your regular expression is valid. Can you show us a little more surrounding code? What do you expect to find in your array when you are done? Mine contains a `1`, because that's what `m//` returns on a successful match. #11929 First ask yourself `How would I do this without a computer?' Then have the computer do it the same way.	[reply] [d/l] [select]
Re^2: Regular expression assistance by agentorange (Sexton) on Dec 07, 2012 at 15:56 UTC
Thanks. Sorry didn't explain it fully but whilst I can pattern match ok I cannot see how to pattern match and pull out the 8th field in the line of which <space> is the separator char.	[reply]
Re^3: Regular expression assistance by kennethk (Abbot) on Dec 07, 2012 at 16:26 UTC
Okay, much better. Assuming you want to do this in the midst of a script and not just piping one liners about, I would use more operations rather than be super sparse on characters. First, since you have a string that is internally-delimited by newlines, you can use `.` to match any character that is not a new line. This means you can actually grab your line using `$cdp =~ m/(.0x01.)/`, storing the line in `$1`. You could then split that on whitespace and grab the 8th field with `(split /\s+/, $1)[7]`. So collecting all your 8th field matches: `my @switch; while ($cdp =~ m/(.0x01.)/g) { push @switch, (split /\s+/, $1)[7]; }` [download] If this is not your intent and the way forward is unclear, post explicit desired output so the goal is clear. #11929 First ask yourself `How would I do this without a computer?' Then have the computer do it the same way.	[reply] [d/l] [select]
Re^4: Regular expression assistance by agentorange (Sexton) on Dec 07, 2012 at 17:01 UTC
Re: Regular expression assistance by jwkrahn (Abbot) on Dec 07, 2012 at 16:06 UTC
VLAN=`echo $cdp \| awk '/0x0a/ { print $8 }'` VLAN=`echo $cdp \| perl -lane'/0x0a/ && print $F[7]'` [download]	[reply] [d/l]
Re: Regular expression assistance by DanEllison (Scribe) on Dec 07, 2012 at 16:08 UTC
Assuming $cdp is piped into perl, try: `while (<>) { if (/Native VLAN ID $0x0a$, length: (\d+) bytes: (\d+)/ +) { print "$2\n"; }}` [download]	[reply] [d/l]


Come for the quick hacks, stay for the epiphanies.
	PerlMonks