[TYPO3-core] RFC: Bug 331: Support IPv6 addresses

[Bernhard Kraft]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Franz Holzinger wrote:

> It is not the task of this function to transform/expand the ip address,
> but to only check the validity.

You can't check if it's valid if you don't expand it.

You do the expansion in normalize IP and I think validIP should do the expansion
and check altogether and return a normalized IP if it's valid and '' if it's invalid
just one more method in t3lib_div and not even one the same as both others :)


I don't know where exactly in your code the bug is but there is one. my test program
wasn't happy with the results your cmpIPv6 returned. I already stripped the test-cases
to only 48 and 64 bit masks (altough I believe when IPv6 will become popular that the
providers will give their customers 112 bit masks and such (giving you one subnet of
65534 possible hosts) - but still masks which have to get expanded fail the cmpIPv6
test.
I think you can't afford from T3 administrators that they write their FULL and exact
ip without and shortcuts like :: or leaving a away preceding 0's ....

so a list like:

192:168::1.0/112,192.168::2.0/112,192.168::3.0/112
should be valid (to allow local subnets 1,2 and 3 access to be) - i know private subnets in
IPv6 are NOT 192:168 .... but that was just an example.


greets,
Bernhard
- --
- ----------------------------------------------------------------------
"Freiheit ist immer auch die Freiheit des Andersdenkenden"
Rosa Luxemburg, 1871 - 1919
- ----------------------------------------------------------------------
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFEDv4QIl4dkVkDMFkRAnvQAKDEez4hn0YGEK6fv1wVSM76yuRxYACguzHq
XB/NM9P0XMAfGr6EutHEGuM=
=w22D
-----END PGP SIGNATURE-----