Last active
March 24, 2016 20:14
-
-
Save anissen/cafce1517bb1994b8003 to your computer and use it in GitHub Desktop.
Parser for simple generative grammar
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
enum Token { | |
TSymbol(s :String); | |
TTerminal(s :String); | |
TArrow; | |
TPlus; | |
TBracketOpen; | |
TBracketClose; | |
TNumber(v :Float); | |
TEof; | |
TEol; | |
} | |
enum Expr { | |
EGenerator(symbol :String, value :Null<Float>, results :Array<String>); | |
} | |
class Lexer extends hxparse.Lexer implements hxparse.RuleBuilder { | |
static public var tok = @:rule [ | |
"=>" => TArrow, | |
"+" => TPlus, | |
"[" => TBracketOpen, | |
"]" => TBracketClose, | |
"[A-Z][a-zA-Z0-9_]*" => TSymbol(lexer.current), // symbol; uppercase starting letter | |
"[a-z][a-zA-Z0-9_]*" => TTerminal(lexer.current), // terminal; lowercase starting letter | |
"(([1-9][0-9]*)|0)(.[0-9]+)?" => TNumber(Std.parseFloat(lexer.current)), | |
"#[^\n\r]*" => lexer.token(tok), // comment | |
"[\t ]" => lexer.token(tok), // whitespace | |
"[\n\r]" => TEol, // line break | |
"" => TEof | |
]; | |
} | |
class Parser extends hxparse.Parser<hxparse.LexerTokenSource<Token>, Token> implements hxparse.ParserBuilder { | |
public function parse() :Array<Expr> { | |
return parseStatements([]); | |
} | |
function parseStatements(stm :Array<Expr>) :Array<Expr> { | |
return switch stream { | |
case [TSymbol(s), v = parseValue(), TArrow, first = parseString(), rest = parseGenerators()]: | |
stm.push(EGenerator(s, v, [first].concat(rest))); | |
parseStatements(stm); | |
case [TEol]: parseStatements(stm); | |
case [TEof]: stm; | |
} | |
} | |
function parseValue() :Null<Float> { | |
return switch stream { | |
case [TBracketOpen, TNumber(v), TBracketClose]: v; | |
case _: null; | |
} | |
} | |
function parseGenerators() :Array<String> { | |
return switch stream { | |
case [TPlus, s = parseString(), e = parseGenerators()]: [s].concat(e); | |
case [TEol]: []; | |
case [TEof]: []; | |
} | |
} | |
function parseString() :String { | |
return switch stream { | |
case [TSymbol(s)]: s; | |
case [TTerminal(s)]: s; | |
} | |
} | |
} | |
class StringEvaluator { | |
static public function eval(e :Expr) :String { | |
return switch(e) { | |
case EGenerator(s, v, r): '$s becomes $r' + (v != null ? ' with probability $v' : ''); | |
} | |
} | |
} | |
class TypeEvaluator { | |
static public function eval(e :Expr) :Dynamic { | |
return switch(e) { | |
case EGenerator(s, v, r): { symbol: s, value: v, results: r }; | |
} | |
} | |
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
class Test { | |
static public function main() { | |
function get_parser(s :String) { | |
var lexer = new GeneratorParser.Lexer(byte.ByteData.ofString(s)); | |
var ts = new hxparse.LexerTokenSource(lexer, GeneratorParser.Lexer.tok); | |
return new GeneratorParser.Parser(ts); | |
} | |
function validate(s :String) { | |
var parser = get_parser(s); | |
try { | |
var parsed :Array<GeneratorParser.Expr> = parser.parse(); | |
for (p in parsed) { | |
var actual = GeneratorParser.StringEvaluator.eval(p); | |
//trace(actual); | |
} | |
} catch (e :hxparse.ParserError) { | |
trace('Parse error', e); | |
trace('==> Validate failure: $s'); | |
} | |
} | |
function invalidate(s :String) { | |
var parser = get_parser(s); | |
try { | |
var parsed :Array<GeneratorParser.Expr> = parser.parse(); | |
trace('==> Invalidate failure: "$s":'); | |
for (p in parsed) { | |
var actual = GeneratorParser.StringEvaluator.eval(p); | |
trace('----> $actual'); | |
} | |
} catch (e :hxparse.ParserError) { | |
// ok | |
} | |
} | |
// Symbol => Symbol | Terminal | |
validate('Symbol => Symbol + terminal'); | |
validate('Symbol => terminal1'); | |
validate('Symbol => terminal_blah1 + terminal2 + terminal3'); | |
trace('------------'); | |
validate('# blah'); | |
validate("Symbol => Symbol1 + Symbol2\nSymbol1 => terminal"); | |
validate('# this is a comment\n\rSymbol => Symbol1 + Symbol2\nSymbol1 => terminal'); | |
trace('------------'); | |
validate('Symbol [20]=> terminal1'); | |
validate('Symbol [12.3]=> terminal1'); | |
validate('Symbol [3.4567989]=> terminal1'); | |
trace('------------'); | |
invalidate('Symbol => Symbol terminal'); | |
invalidate('Symbol => Symbol ++ terminal'); | |
invalidate('Symbol =>=> term'); | |
} | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment