Comparing db9d5fff5e..9eb2d8c5bd - groove/routeros-scripts-main - Forgejo: Beyond coding. We Forge.

groove/routeros-scripts-main

mirror of https://github.com/eworm-de/routeros-scripts.git synced 2026-03-06 14:09:35 +00:00

Author	SHA1	Message	Date
Christian Hesse	9eb2d8c5bd	Merge branch 'netmask6' into next This branch is a follow-up on `9ceed0926a` with clean solution. Read on for details... The data type `ip6-prefix` used to hold what it was named for - an IPv6 prefix: [user@mikrotik] > :put 2001:db8::dead:beef/32 2001:db8::/32 This changed with RouterOS 6.21beta2, which now allows that exact same data type to hold something like "address with prefix length attached": [user@mikrotik] > :put 2001:db8::dead:beef/32 2001:db8::dead:beef/32 My scripts (namely `fw-addr-lists`) relied on the old behaviour and broke. The commit mentioned above was just a quick workaround, with rough edges, and it could still fail. Sadly RouterOS does not support bit shifting on IPv6 data types, so a (completely) mathematical solution is out of scope. This branch implements a new and better workaround, see the first commit of branch for details. I opened a support ticket / feature request on this topic, let's see what results it brings... https://help.mikrotik.com/servicedesk/servicedesk/customer/portal/1/SUP-201881	2025-10-19 20:00:38 +02:00
Christian Hesse	b80b872e55	mod/ipcalc: support IPv6 Well, some of these values do not make a lot of sense for IPv6... Something to be cleaned up later.	2025-10-19 19:55:12 +02:00
Christian Hesse	ea05b69f7c	fw-addr-lists: use $NetMask6	2025-10-19 19:55:12 +02:00
Christian Hesse	d7a6eb1d00	global-functions: $NetMask6: implement simple caching The calculation is quite complex for something that needs to be done frequently, for example by `fw-addr-lists`. The number of possible netmasks is limited, so let's cache the results that were calculated already.	2025-10-19 19:52:42 +02:00
Christian Hesse	6ad6f9aa08	global-functions: introduce $NetMask6 RouterOS does not support bit shifting on IPv6 data types, so we have to split the problem: * each 16 bit block is calculated separately, as number * the complete netmask is assembled in a loop, as string * the final string is casted to correct data type	2025-10-19 19:48:14 +02:00
Christian Hesse	c62f236251	Merge branch 'netmask4' into next	2025-10-19 19:26:51 +02:00

Diff content is not available