DSL: Expand to NWC algorithms

[anshumanmohan]

We are now fully committing to supporting non work-conserving algorithms, so let's flesh the DSL out appropriately.

• [x] Research and document a few new NWC algorithms in the policies glossary.
• [x] Extend the AST to include a few NWC algorithms.
• [x] Extend the parser appropriately.
• [x] Extend the suite of example programs to include NWC algorithms.

[KabirSamsi]

Working on this with reference (and conjunction) to #13. I think a good approach is to extend the AST to include the non-work-conserving policies outlined there, and as we find other interesting ones, we can update both.

[anshumanmohan]

Sounds good! But just a heads up, there are no NWC policies in the glossary at present. If you're taking this trio of tasks on, you'll actually need to first do a 0th task, which is to review some scheduling papers (I recommend starting with good old Sivaraman '16) to learn about some NWC policies, and to them add those to the glossary.

[KabirSamsi]

My mistake! I'll put that together All the same, I think it's definitely worthwhile to add the remaining existing work-conserving policies from the glossary into the DSL, especially since it hasn't been too much work.

[anshumanmohan]

Yes that sounds great! And, just to make connections across conversations, these would all still be "simple" policies as defined in https://github.com/cucapra/packet-scheduling/discussions/38. While you're at it, maybe rename Fair in the DSL into RoundRobin for consistency with the glossary?

[KabirSamsi]

13 is now expanded to include four new, non-work-conserving policies – RCSP, Leaky Bucket, Token Bucket and Stop-And-Go. The latter three of these form a paradigm of non-work-conserving policies called 'traffic control policies' – essentially, ones which, rather than providing new levels of expressivity, primarily function to regulate the nature of the flow of packets and transactions.

The AST now includes these policies, but more metadata is required for the newest non-conserving ones.