gll
gll copied to clipboard
rust-lang
GLL parsing framework.
The `indexing` crate already allows appending to the container (that we use as our input), the main thing missing is the ability to yield "not ready" for a specific thread...
Allow supplying one (or more?) range of changes to a past input, and reuse the memoizer and SPPF. Maybe we can use a non-contiguous range space? That way we instead...
See https://github.com/rust-lang/rust/pull/52650 - the main reason there's not a trait right now is because the `all` method returns `-> impl Trait`, and there wasn't a way of doing that in...
This can be done with ref-counting, but to avoid using an actual `Rc` (e.g. to be able to do #9) and/or the GSS being more tightly coupled with the SPPF,...
#### GLL * https://www.sciencedirect.com/science/article/pii/S1571066110001209 * https://www.sciencedirect.com/science/article/pii/S0167642312000627 * https://cdn.rawgit.com/iguana-parser/papers/master/cc15.pdf * https://speakerdeck.com/afroozeh/faster-practical-gll-parsing * https://www.sciencedirect.com/science/article/pii/S016764231630003X * https://link.springer.com/chapter/10.1007%2F978-3-319-71734-0_3 * https://pure.tue.nl/ws/files/46987757/782897-1.pdf #### SPPF * http://www.bramvandersanden.com/post/2014/06/shared-packed-parse-forest/ * http://ceur-ws.org/Vol-1403/paper2.pdf * https://people.csail.mit.edu/jgross/personal-website/papers/2015-parsing-parse-trees.pdf * https://pure.tue.nl/ws/files/46998704/784691-1.pdf #### RCG * http://www.sfs.uni-tuebingen.de/emmy/esslli/slides/4rcg.pdf...
Because of how GLL control-flow maps all the way down to machine code, it should be possible for a program-trace-aware fuzzer to hit enough codepaths to cause an ambiguity through...
@eternaleye keeps bringing them up and I keep forgetting they exist.
If we can encode an unambiguous (parse) subtree (from #40) in Pola's types, we might be able to provide a more restricted language, to write disambiguating logic in, based on...
Right now, `Handle` only knows the range `X` matched at, which always presents the same view of `X`'s substructure (i.e. the same ambiguities are always observed when visiting its descendants)....
Blocked on #33. Also, not *entirely* sure this is a good idea.