[Q] How to hook up the GUILE parser?

[Han-Wen Nienhuys <hanwen at cs dot uu dot nl>]

Hi guys,

We're still advancing GUILE integration in LilyPond. For this purpose,
I want to parse arbitrary Scheme expressions that are embedded in our
input. I want the expressions to be embedded like this:

       \notes {
         c4
	   \property Voice.stemSettings = #(map sqr '(1 2 3))
	     c4
	     }

My idea was have a special character (the '#') for switching
the lexer in a separate mode; The rough sketch of my  idea is this:

* Tokenizer finds a #.  Tokenizer sets up a string-port (using the
  input file, which completely in core), that starts where the
  tokenizer found the # in the file

* Tokenizer calls read-and-eval! with said string-port.
  read-and-eval!
  processes one expression and returns the result.

* Tokenizer reads how many characters were read from the string-port

* Tokenizer switches into special mode `guzzle-characters'
  and gulps as many characters from the input as were read from the
  port.

* Tokenizer signals to parser "found token of type SCM".

Now my question: how do I set up a string port that does this, and how
do I read back the number of characters taken from the port?

Any other suggestions?

BTW, please CC replies, I'm not on the GUILE list.


-- 

Han-Wen Nienhuys, hanwen@cs.uu.nl ** GNU LilyPond - The Music Typesetter 
      http://www.cs.uu.nl/people/hanwen/lilypond/index.html