如何从网络中删除两个或多个子网？

Question

from ipaddress class, I know the address_exclude method. below is an example from the documentation:

>>> n1 = ip_network('192.0.2.0/28')
>>> n2 = ip_network('192.0.2.1/32')
>>> list(n1.address_exclude(n2))  
[IPv4Network('192.0.2.8/29'), IPv4Network('192.0.2.4/30'),
 IPv4Network('192.0.2.2/31'), IPv4Network('192.0.2.0/32')]

but what about if I want to remove two or more subnets from a network? for example, how can I delete from the 192.168.10.0/26 his subnets 192.168.10.24/29 and 192.168.10.48/28? the result should be 192.168.10.0/28, 192.168.10.16/29 and 192.168.10.32/28.

I'm trying to find a way to write the algoritm that I use in my mind using the address_exclude method but I can't. is there a simple way to implement what I just explained?

Answer 1

When you exclude one network from another, the result can be multiple networks (original one got split) - so, for the rest of the networks to exclude, you need to first find which part they would fit into before excluding them as well.

Here's one possible solution:

from ipaddress import ip_network, collapse_addresses

complete = ip_network('192.168.10.0/26')

# I chose the larger subnet for exclusion first, can be automated with network comparison
subnets = list(complete.address_exclude(ip_network('192.168.10.48/28')))
# other network to exclude
other_exclude = ip_network('192.168.10.24/29')

result = []
# Find which subnet the other exclusion will happen in 
for sub in subnets:
    # If found, exclude & add the result
    if other_exclude.subnet_of(sub):
        result.extend(list(sub.address_exclude(other_exclude)))
    else:
    # Other subnets can be added directly
        result.append(sub)

# Collapse in case of overlaps
print(list(collapse_addresses(result)))

Output:

[IPv4Network('192.168.10.0/28'), IPv4Network('192.168.10.16/29'), IPv4Network('192.168.10.32/28')]

Answer 2

Expanding on my brain wave posted on @rdas's response, posting my solution.

It seems better to split the initial network into the smallest chunks you are asking, and do this for all ranges to be removed. Then exclude them from the list and return result.

from ipaddress import ip_network, collapse_addresses

def remove_ranges(mynetwork,l_of_ranges):
    # find smallest chunk
    l_chunk = sorted(list(set([x.split('/')[1] for x in l_of_ranges])))
    l_mynetwork = list(ip_network(mynetwork).subnets(new_prefix=int(l_chunk[-1])))

    l_chunked_ranges = [ ]
    for nw in l_of_ranges:
        l_chunked_ranges.extend(list(ip_network(nw).subnets(new_prefix=int(l_chunk[-1]))))

    l_removed_networks = [ ]
    for mynw in l_mynetwork:
        if not mynw in l_chunked_ranges:
            l_removed_networks.append(mynw)

    result = list(collapse_addresses(l_removed_networks))

    return [str(r) for r in result]


if __name__ == '__main__':
    mynetwork = "10.110.0.0/16"
    l_of_ranges = ["10.110.0.0/18","10.110.72.0/21","10.110.80.0/21","10.110.96.0/21"]

    print(f"My network: {mynetwork}, Existing: {l_of_ranges} ")
    a = remove_ranges(mynetwork,l_of_ranges)
    print(f"Remaining: {a}")

With the result:

My network: 10.110.0.0/16, Existing: ['10.110.0.0/18', '10.110.72.0/21', '10.110.80.0/21', '10.110.96.0/21'] 
Remaining: ['10.110.64.0/21', '10.110.88.0/21', '10.110.104.0/21', '10.110.112.0/20', '10.110.128.0/17']

Which seems to be valid.