[PATCH] D127357: [pseudo] wip/prototype: use LR0 instead of SLR1 table

[Sam McCall via Phabricator via cfe-commits]

sammccall added a comment.

In D127357#3569386 <https://reviews.llvm.org/D127357#3569386>, @hokein wrote:

> Thanks for experimenting this!
>
>> the size of the LR table is much smaller (In this patch we go from 340kB => 120kB, and the encoding isn't efficient)
>
> This is interesting. I'd expect we now add more reduce actions into the LRTable, thus the table size should grow (I guess the saving is because we don't need to store different token kinds, instead just a single `eod` token)?

Yes, before we store (State, Tok) => Reduce(L := R) for every Tok in Follow(L).
Now we only store (State, eod) => Reduce(L := R).


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D127357/new/

https://reviews.llvm.org/D127357