Narsese Lexer and Parser

[PtrMan]

first partially done version.

untested (due to the usual git chaos in GUI)

[patham9]

Why does the core need a Lexer? The PR needs to be made much more convincing, as I can't really suggest changing the working parser before we are done with MVP1.

[PtrMan]

Why not MVP0 - I wouldn't call it working - it is semi-working

[PtrMan]

Yes because I need this for the rule language (or a approximation of it) and syntax highlighting in GUI

[patham9]

because MVP0 should not have experimental stuff. But feel free to add it in nars_lab. I don't mind ideas getting developed incrementally there, which, when mature, eventually make it into nars_core. And personally, as you know, I see the adding of a new parser as mostly a waste of time to make software-philes happy, but I won't dictate how you should spend your time. But also consider that a lot of invalid statements can only be detected after the parsing, after reduction and normalization, and this is what is already happening to a large extent.

[PtrMan]

That are different responsibilities - and I am aware of it.

[PtrMan]

This can't be done fully in labs (which is a good idea) because the parser can't be overwritten