Strange net paths in WHOIS

As a self-learning project, I'm implementing a C++ whois client. I'm using a couple of ancient C implementations for reference, e.g. koders.com.

One thing puzzles me. As well as the expected path format, like example.com, example.co.uk, 192.0.32.10 or 2620:0:2d0:200::10, the old C implementations accept some other formats:

1. Paths beginning with one of the following strings: "net-", "netblk-", "asn-", "as-", "lim-", "coco-", "coho-" or "core-".

I presume these look something like "netblk-example" but I don't have a working example.

1. Paths ending in one of the following strings: "-au-dom", "-dom", "-org", "-hst", "-arin", "-ripe", "-mnt", "-gandi", "-ap", "-au", "-ti", "-is", "-6bone", "-norid", "-ripn", "-sgnic", "-metu", "-cknic" or "-kg".

Again, I presume these look like "example-arin" but I'm not certain.

To (finally) get to my question: does anybody know what these paths are? Are they still used? Does anybody have legit examples of valid paths like these?

Update 23 May 2011 Added C++ tag (my implementation language) as nobody seems to even开发者_JS百科 view "network-programming" without a language.

---

These extra strings are (mostly) for recognising the object types maintained by the Regional Internet Registries, such as RIPE, ARIN, APNIC.

Their databases are also accessible via the whois protocol, even though the entries within them are not domain names.