(* wmap Language Specification *)

(* Top-level structure *)
program = { line } ;
line = ( entity line_ending ) | ( ignored_line line_ending ) | line_ending ;

entity = component | dependency | note | stage | group | inertia | evolution ;
ignored_line = ? any sequence of tokens that doesn't match entity grammar ? ;

(* Entities *)
component = component_label position [ shape ] ;
dependency = component_label dependency_type component_label ;
note = case_insensitive("[Note]") position text ;
stage = "[" stage_number "]" real_number ;
group = case_insensitive("[Group]") component_label { "," component_label } ;
inertia = case_insensitive("[Inertia]") component_label ;
evolution = case_insensitive("[Evolution]") component_label sign real_number ;

(* Component definitions *)
position = "(" real_number "," real_number ")" ;
shape = "[" shape_label "]" ;
dependency_type = "--" | "->" ;
sign = "+" | "-" ;

(* Terminal symbols *)
component_label = preserved_case( { character - ( "-" | "+" | "," | "[" | "]" | "(" | ")" | line_ending ) } ) ;
text = preserved_case( { character - line_ending } ) ;
real_number = [ digit ] { digit } [ "." { digit } ] | "." digit { digit } ;
stage_number = case_insensitive( "i" | "ii" | "iii" | "iv" ) ;
shape_label = case_insensitive( "x" | "square" | "triangle" | "circle" ) ;

(* Lexical elements *)
character = ? any Unicode character ? ;
digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9" ;
line_ending = lf | crlf | cr ;
lf = ? U+000A (Line Feed) ? ;
cr = ? U+000D (Carriage Return) ? ;
crlf = cr lf ;

(* Lexical rules *)
case_insensitive(x) = ? case-insensitive match of x ? ;
preserved_case(x) = ? case-insensitive match of x, preserving original case for display ? ;

(* Notes *)
(* - Keywords and shape labels are case-insensitive *)
(* - Component labels are case-insensitive for matching but preserve original case *)
(* - Text content preserves original case *)
(* - Whitespace (except line endings) may appear between any tokens *)
(* - Each line contains exactly one entity *)
(* - Invalid lines should be ignored *)
(* - Line endings support Unix (LF), Windows (CRLF), and classic Mac (CR) formats *)
(* - CRLF is treated as a single line ending, not separate CR and LF *)