Hacking S-expressions into Ruby

#1

Post by sonnym » Fri, Feb 27, 2015, 1:31 PM UTC

Hacking S-expressions into Ruby
sonnym.github.io

Re: Hacking S-expressions into Ruby

#2

Post by _mikz » Fri, Feb 27, 2015, 6:51 PM UTC

` puts RubyVM::InstructionSequence.compile('[]').disasm` points to `newarray` which is YARV instruction: https://github.com/ruby/ruby/blob/ca6b174078fa15f33655be704d...

Re: Hacking S-expressions into Ruby

#3

Post by lectrick » Fri, Feb 27, 2015, 7:22 PM UTC

_mikz wrote: post #2

` puts RubyVM::InstructionSequence.compile('[]').disasm` points to `newarray` which is YARV instruction: https://github.com/ruby/ruby/blob/ca6b174078fa15f33655be704d...

I don't suppose there's a way to hook into newarray ...?

Maybe a gem with a C component could provide a hook?

Re: Hacking S-expressions into Ruby

#4

Post by aaronem » Fri, Feb 27, 2015, 7:46 PM UTC

lectrick wrote: post #3

_mikz wrote: post #2

` puts RubyVM::InstructionSequence.compile('[]').disasm` points to `newarray` which is YARV instruction: https://github.com/ruby/ruby/blob/ca6b174078fa15f33655be704d...

I don't suppose there's a way to hook into newarray ...? Maybe a gem with a C component could provide a hook?

At some point it's surely simpler just to write a Lisp interpreter instead.

Re: Hacking S-expressions into Ruby

#5

Post by lookatparsetree » Fri, Feb 27, 2015, 10:16 PM UTC

Have a look at https://github.com/seattlerb/parsetree

Re: Hacking S-expressions into Ruby

#6

Post by bradleyland » Sat, Feb 28, 2015, 4:31 PM UTC

lookatparsetree wrote: post #5

Have a look at https://github.com/seattlerb/parsetree

That version only works with 1.8. There is an updated fork though, sexp_processor:

https://github.com/seattlerb/sexp_processor

Re: Hacking S-expressions into Ruby

#7

Post by lectrick » Mon, Mar 02, 2015, 12:43 PM UTC

aaronem wrote: post #4

lectrick wrote: post #3

Earlier quoted context omitted.

I don't suppose there's a way to hook into newarray ...? Maybe a gem with a C component could provide a hook?

At some point it's surely simpler just to write a Lisp interpreter instead.

Or switch to Elixir and its nifty macro system, which I did.