Validating a subnet mask

I'm working on some code that handles IPs entered as strings and stored as arraylists of integers. That part of the code is up and running.

Now I want to be able to enter subnet masks. This re-uses most of the code, but requires some extra validation (since only a few addresses are valid as subnet masks - see here). However, I'm not sure how to handle this since subnet masks follow some very specific rules.

Plan B is to just check against all valid subnet masks but that seems clumsy. Has anyone else had this problem before? Can anyone point me in the right direction to solve this?

Wouter Hermans wrote: However, I'm not sure how to handle this since subnet masks follow some very specific rules.

Take a look at all the valid submasks -- but this time examine it in binary. You will see that it follows a very specific pattern. And you can easily write a routine to validate it.

Henry

Instead of reusing the code, can't you extend / encapsulate the existing validation class(es)?

Rob Spoor wrote:Instead of reusing the code, can't you extend / encapsulate the existing validation class(es)?

(Sorry for the late reply)

The subnet mask code extends my IP class and reuses the bulk of its code, but the rules for subnets are slightly different.

As for the binary suggestion, that definitely seems like a good idea. Thanks.

Wouter Hermans wrote:

Rob Spoor wrote:Instead of reusing the code, can't you extend / encapsulate the existing validation class(es)?

The subnet mask code extends my IP class and reuses the bulk of its code, but the rules for subnets are slightly different.

I know. Basically you already followed my advice before I even gave it. I just thought that when you said "re-uses most of the code" you copy-pasted the code, but you actually meant inherited the methods and their behaviour.