[NTLUG:Discuss] Is there an easy way?

[Leroy Tennison]

Carl Haddick wrote:
> On Sun, Jul 26, 2009 at 12:06:14AM -0500, Leroy Tennison wrote:
>> I'm not even sure how to Google for this one without having to wade 
>> through too many extraneous hits so I'm posting here.  I have three 
>> files and need to merge fields from each of them into a single file.  A 
>> brief description of each is:
>>
>> File 1 has a phone number and an associated IP address (VOIP, Avaya PBX)
>> File 2 has an IP address and an associated MAC address (Cisco router).
>> File 3 has a MAC address and associated Cisco switch port assignment.
>>
>> I need a file containing phone number and associated switch port.
>>
>> This may be somewhat OT except I'm hoping there's a Linux utility which 
>> will meet the need.  I know that I can turn each file into a database 
>> table and do some operation on two tables to "merge" their fields and 
>> then repeat again to get what I need but that's the extent of my 
>> understanding of that option.  Wondering if there's a utility to take 
>> the files directly and do the "merge".  This is a situation where 
>> scripting seems to be the wrong solution, I've written something similar 
>> in awk and don't want to do that again.  I'm also not looking forward to 
>> writing a program in a "real programming language" to accomplish this 
>> either.  Thanks for any input.
>>
>> _______________________________________________
>> http://www.ntlug.org/mailman/listinfo/discuss
> 
> Ok, I'll admit, I keep hoping someday somebody will need a Python
> hacker.  I guess I really know how to pick languages, I've also got
> extensive Forth experience.  But I gotta say for some things Forth can
> really get the job done.  If I was ever stranded on a desert island with
> nothing but a PDP11 running nothing but a front panel I'd have Forth
> running in a week, which would let me write a simple word processor in
> another week, and then I could print out my message to tuck in a handy
> bottle...
> 
> Anyway, two 'for' loops in Python came to mind:
> 
> #!/usr/local/bin/python
> 
> phoneip={}
> macswitch={}
> ipmac={}
> 
> for d in zip(
> 	(file('phoneip').readlines(),
> 	    file('ipmac').readlines(),
> 	    file('macswitch').readlines()),
> 	(phoneip,
> 	    ipmac,
> 	    macswitch)
> 	):
>     for l in d[0]:
> 	key,val=l.strip().rsplit(' ',1)
> 	d[1][key]=val
> 
> for p in phoneip:
>     ip=phoneip[p]
>     mac=ipmac[ip]
>     port=macswitch[mac]
>     print 'Phone:%s has ip: %s, mac: %s, and is on port: %s'%(p,ip,mac,port)
> 
> The output from that is like:
> 
> Phone:(666) 234-5678 has ip: 4.5.6.7, mac: 22:33:44:55:66:77:88:99, and
> is on port: p2
> Phone:(777) 345-6789 has ip: 7.8.9.10, mac: 33:44:55:66:77:88:aa, and is
> on port: p3
> Phone:(555) 123-4567 has ip: 1.2.3.4, mac: 11:22:33:44:55:66:77:88, and
> is on port: p1
> 
> But don't let me argue against the simple approach.  Cut, join, uniq,
> sort, and most certainly awk are all supremely useful.  However, if you
> need to pick through the files, regex out what you need, maybe do some
> other housekeeping, a script approach might be the easy way out.
> 
> Three for loops.  That was three for loops in my Python blatherings, not
> two.
> 
> Carl
> 
> _______________________________________________
> http://www.ntlug.org/mailman/listinfo/discuss
> 
Thanks for the input, I found that join really did meet the need.