Proposal: Introduce some hook to change encoding of parsed result about spitfire HOT 3 OPEN

@mhanberg Zero worries! This is low prio. Feel better.

from spitfire.

@zachallaun I think the premise is reasonable, not entirely sure how it would look off the top of my head.

Naively, I think you could wrap every line that constructs a node with a function that invokes either the node_encoder or does the default behavior. There are probably some places where the AST is being changed in different places after its constructed, which might make it difficult.

Please submit a spike whenever you get some time, it would be helpful to see what it might look like. It's entirely possible that a pluggable encoder still won't be able to achieve the desire you have, so it'd be good to spike out a more intricate case.

As a concrete example: when parsing Foo.Bar.\nBaz (note the newline), it's not possible to determine which line Bar occurs on without inspecting the source, but with token data, it would be.

Side note, this in particular could potentially be upstreamed to the core yecc parser, and then I could mimic the behavior into Spitfire.

from spitfire.

Hi, sorry I've been off the computer for a while due to a back issue, just getting back in.

I'll give this a thorough read this week and reply.

from spitfire.