4.7 BNF Grammars
The scribble/bnf library
provides utilities for typesetting grammars.
For example,
@(let ([open @litchar{(}] |
[close @litchar{)}]) |
@BNF[(list @nonterm{expr} |
@nonterm{id} |
@BNF-seq[open @kleeneplus[@nonterm{expr}] close] |
@BNF-seq[open @litchar{lambda} |
open @kleenestar[@nonterm{id}] close |
@nonterm{expr} close] |
@nonterm{val}) |
(list @nonterm{val} |
@BNF-alt[@nonterm{number} @nonterm{primop}]) |
(list @nonterm{id} |
@elem{any name except for @litchar{lambda}})]) |
produces the output
| ‹expr› | ::= | ‹id› |
|
| | | ( ‹expr›+ ) |
|
| | | ( lambda ( ‹id›* ) ‹expr› ) |
|
| | | ‹val› |
| ‹val› | ::= | ‹number› | ‹primop› |
| ‹id› | ::= | any name except for lambda |
See also racketgrammar.
Typesets a grammar table. Each production starts with an element
(typically constructed with nonterm) for the non-terminal
being defined, and then a list of possibilities (typically constructed
with BNF-seq, etc.) to show on separate lines.
Typesets a non-terminal: italic in angle brackets.
Typesets a sequence.
Typesets a group surrounded by curly braces (so the entire group can
be repeated, for example).
Typesets an optional element: in square brackets.
(kleenestar pre-content ...) → element? pre-content : any/c
Typesets a 0-or-more repetition.
(kleeneplus pre-content ...) → element? pre-content : any/c
Typesets a 1-or-more repetition.
(kleenerange n m pre-content ...) → element? n : any/c m : any/c pre-content : any/c
Typesets a n-to-m repetition. The n and
m arguments are converted to a string using (format "~a" n) and (format "~a" m).
Typesets alternatives for a production’s right-hand side to appear on
a single line. The result is normally used as a single possibility in
a production list for BNF.
A string to use for omitted productions or content.