Pag Grammar Definition

This is a work in Progress

You read that right, this is a grammar for our grammar. That is, this defines the format for the grammars one can use for procedural generation.

PAG := SECTION⁺

SECTION := COMMENT NEWLINE SECTION | PRIORITY NEWLINE SECTION | RULE

COMMENT := ##[^NEWLINE]⁺

PRIORITY := Priority NUMBER:

RULE := (PREDECESSOR : CONDITIONAL -> SUCCESSOR : PROBABILITY)

PREDECESSOR := ALPHANUMERIC⁺(LPAREN PARAM(,PARAM)^* RPAREN)?

CONDITIONAL := CONDVAR OP CONDVAR | OCCLUSION | !LPAREN CONDITIONAL RPAREN | CONDITIONAL && CONDITIONAL | CONDITIONAL || CONDITIONAL

CONDVAR := PROBVAR | VARARITHMETIC

SCOPEVAR := Scope\.(X | Y | Z | S | P)

OCCLUSION := (Shape.occ LPAREN "LETTER⁺ " RPAREN (== | !=) ("none" | "part" | "full")) | (Shape.visible LPAREN "LETTER⁺ " RPAREN)

OP := == | != | > | >= | < | <=

PROBABILITY := FLOAT | VARARITHMETIC

VARARITHMETIC := PROBVAR ARITHMETIC PROBVAR | LPAREN VARARITHMETIC RPAREN ARITHMETIC LPAREN VARARITHMETIC RPAREN

PROBVAR := SCOPEVAR | NUMBER | PARAM | FLOAT

ARITHMETIC := \+ | - | \* | \\

SUCCESSOR := PREDECESSOR | BASERULE

BASERULE := TODO

PARAM := LETTER

ALPHANUMERIC := LETTER | NUMBER

NUMBER := [0-9]

LETTER := [A-Za-z]

LPAREN := \ (

RPAREN := \ )

FLOAT := NUMBER^*\.NUMBER⁺

NEWLINE := \n (maybe use \s instead?)

I use the regular expression notation for + (1 or more), ? (at most 1), brackets for characters, * (for 0 or more), backslash for escaping, and parenthesis for grouping.