luaj/grammar/Lua52.jj

/** 
* Javacc grammar for lua language version 5.2
*
* Originally created for use in luaj, a Java implementation of the lua language
* @see http://sourceforge.net/projects/luaj/
* 
* For documentation on the lua language 
* @see http://www.lua.org/manual/5.2/
*
* Empty grammar that validates syntax without producing a parse tree.  
*
* @author Jim Roseborough
* @date August 30, 2012
*/

options {
  STATIC = false;
  JDK_VERSION = "1.3";
  ERROR_REPORTING = false;
  DEBUG_LOOKAHEAD = false;
  DEBUG_PARSER = false;
  DEBUG_TOKEN_MANAGER = false;
  OUTPUT_DIRECTORY = "org/luaj/vm2/parser";
}

PARSER_BEGIN(LuaParser)
package org.luaj.vm2.parser;

public class LuaParser {

  public static void main(String args[]) throws ParseException {
    LuaParser parser = new LuaParser(System.in);
    parser.Chunk();
  }

  public static final int VAR  = 0;
  public static final int CALL = 1;
}

PARSER_END(LuaParser)

/* WHITE SPACE */

SKIP :
{
  " " | "\t" | "\n" | "\r" | "\f"
}

/* COMMENTS and LONG STRINGS */

MORE :
{
  "--[[": IN_LC0
| "--[=[": IN_LC1
| "--[==[": IN_LC2
| "--[===[": IN_LC3
| < "--[====" ("=")* "[" > : IN_LCN
| "[[" : IN_LS0
| "[=[" : IN_LS1
| "[==[" : IN_LS2
| "[===[" : IN_LS3
| < "[====" ("=")* "[" > : IN_LSN
| "--"   : IN_COMMENT
}

<IN_COMMENT> SPECIAL_TOKEN :
{
  <COMMENT: (~["\n","\r"])* ("\n"|"\r"|"\r\n")? > : DEFAULT
}

<IN_LC0> SPECIAL_TOKEN : { <LONGCOMMENT0: "]]" > : DEFAULT }
<IN_LC1> SPECIAL_TOKEN : { <LONGCOMMENT1: "]=]" > : DEFAULT }
<IN_LC2> SPECIAL_TOKEN : { <LONGCOMMENT2: "]==]" > : DEFAULT }
<IN_LC3> SPECIAL_TOKEN : { <LONGCOMMENT3: "]===]" > : DEFAULT }
<IN_LCN> SPECIAL_TOKEN : { <LONGCOMMENTN: "]====" ("=")* "]" > : DEFAULT }

<IN_LS0> TOKEN : { <LONGSTRING0: "]]" > : DEFAULT }
<IN_LS1> TOKEN : { <LONGSTRING1: "]=]" > : DEFAULT }
<IN_LS2> TOKEN : { <LONGSTRING2: "]==]" > : DEFAULT }
<IN_LS3> TOKEN : { <LONGSTRING3: "]===]" > : DEFAULT }
<IN_LSN> TOKEN : { <LONGSTRINGN: "]====" ("=")* "]" > : DEFAULT }

<IN_LC0,IN_LC1,IN_LC2,IN_LC3,IN_LCN,IN_LS0,IN_LS1,IN_LS2,IN_LS3,IN_LSN> MORE :
{
  < ~[] >
}


/* RESERVED WORDS AND LITERALS */

TOKEN :
{
  <AND: "and">
| <BREAK: "break">
| <DO: "do">
| <ELSE: "else">
| <ELSEIF: "elseif">
| <END: "end">
| <FALSE: "false">
| <FOR: "for">
| <FUNCTION: "function">
| <GOTO: "goto">
| <IF: "if">
| <IN: "in">
| <LOCAL: "local">
| <NIL: "nil">
| <NOT: "not">
| <OR: "or">
| <RETURN: "return">
| <REPEAT: "repeat">
| <THEN: "then">
| <TRUE: "true">
| <UNTIL: "until">
| <WHILE: "while">
}

/* LITERALS */

TOKEN :
{
  < NAME: ["a"-"z", "A"-"Z", "_"] (["a"-"z", "A"-"Z", "_", "0"-"9"])* >
| < NUMBER: <HEX> | <FLOAT> >
| < #FLOAT: <FNUM> (<EXP>)? >
| < #FNUM: (<DIGIT>)+ "." (<DIGIT>)* | "." (<DIGIT>)+ | (<DIGIT>)+ >
| < #DIGIT: ["0"-"9"] >
| < #EXP: ["e","E"] (["+","-"])? (<DIGIT>)+ >
| < #HEX: "0" ["x","X"] <HEXNUM> (<HEXEXP>)? >
| < #HEXNUM: (<HEXDIGIT>)+ "." (<HEXDIGIT>)* | "." (<HEXDIGIT>)+ | (<HEXDIGIT>)+ >
| < #HEXDIGIT: ["0"-"9","a"-"f","A"-"F"] >
| < #HEXEXP: ["e","E","p","P"] (["+","-"])? (<DIGIT>)+ >
| < STRING: "\"" (<QUOTED> | ~["\\","\""])* "\"" >
| < CHARSTRING: "'" (<QUOTED> | ~["\\","'"])* "'" >
| < #QUOTED: <DECIMAL> | <UNICODE> | <CHAR> >
| < #DECIMAL:   "\\" ["0"-"9"] (["0"-"9"])? (["0"-"9"])? >
| < DBCOLON: "::" >
| < #UNICODE: "\\" "u" <HEXDIGIT> <HEXDIGIT> <HEXDIGIT> <HEXDIGIT> >
| < #CHAR: "\\" (~[]) >
| < #LF: ("\n" | "\r" | "\r\n") >
}

/** Root production. */
void Chunk():
{}
{
	( "#" { token_source.SwitchTo(IN_COMMENT); } )? Block() <EOF>
}

void Block():
{}
{
	( Stat() )* ( ReturnStat() )? 
}

void Stat():
{}
{
	";"
|	Label()
|	<BREAK>
|	<GOTO> <NAME>
|	<DO> Block() <END>
|	<WHILE> Exp() <DO> Block() <END>
|	<REPEAT> Block() <UNTIL> Exp() 
|	<IF> Exp() <THEN> Block() (<ELSEIF> Exp() <THEN> Block())* (<ELSE> Block())? <END>
|	LOOKAHEAD(3) <FOR> <NAME> "=" Exp() "," Exp() ( "," Exp() )? <DO> Block() <END>
|	<FOR> NameList() <IN> ExpList() <DO> Block() <END>
|	<FUNCTION> FuncName() FuncBody()
|	LOOKAHEAD(2) <LOCAL> <FUNCTION> <NAME> FuncBody()
|	<LOCAL> NameList() ( "=" ExpList() )? 
|	ExprStat()
}

void ReturnStat():
{}
{
	<RETURN> ( ExpList() )? ( ";" )?
}

void Label():
{}
{
	<DBCOLON> <NAME> <DBCOLON>
}

void ExprStat():
{	int type,need=CALL; }
{
	type=PrimaryExp() ( Assign() { need=VAR; } )?
	{ if ( type!=need ) throw new ParseException("expected function call or assignment"); }
}

void Assign():
{}
{
	( "," VarExp() )* "=" ExpList()
}

void VarExp():
{	int type; }
{
	type=PrimaryExp()
	{ if ( type!=VAR ) throw new ParseException("expected variable expression"); }
}

void FuncName():
{}
{
	<NAME> ( "." <NAME> )* ( ":" <NAME> )?
}

void PrefixExp():
{}
{
	<NAME>
|	ParenExp()
}

void ParenExp():
{}
{
	"(" Exp() ")"
}

int PrimaryExp(): 
{	int type=VAR; }
{
	PrefixExp() ( LOOKAHEAD(2) type=PostfixOp() )* { return type; }
}

int PostfixOp():
{}
{
	FieldOp() { return VAR; }
|	FuncOp()  { return CALL;  }
}

void FieldOp():
{}
{
	"." <NAME> 
|	"[" Exp() "]"
}

void FuncOp():
{}
{
	":" <NAME> FuncArgs() 
|	FuncArgs() 
}

void FuncArgs():
{}
{
	"(" ( ExpList() )? ")"
|	TableConstructor()
|	Str()
}

void NameList():
{}
{
	<NAME> ( LOOKAHEAD(2) "," <NAME> )*
}

void ExpList():
{}
{
	Exp() ( "," Exp() )*
}

void SimpleExp():
{}
{
	<NIL> 
|	<TRUE> 
|	<FALSE>
|	<NUMBER>
|	Str()
|	"..." 
|	TableConstructor()
|	FunctionCall() 
|	PrimaryExp() 
}

void Str():
{}
{
	<STRING> 
|	<CHARSTRING>
|	<LONGSTRING0>
|	<LONGSTRING1>
|	<LONGSTRING2>
|	<LONGSTRING3>
|	<LONGSTRINGN>
}

void Exp():
{}
{
	SubExp()
}

void SubExp():
{}
{
	( SimpleExp() | Unop() SubExp() ) (LOOKAHEAD(2) Binop() SubExp())*
}

void FunctionCall():
{}
{
	<FUNCTION> FuncBody()
}

void FuncBody():
{}
{
	"(" ( ParList() )? ")" Block() <END>
}

void ParList():
{}
{
	NameList() ( "," "..." )? | "..." 
}

void TableConstructor():
{}
{
	"{" ( FieldList() )? "}"
}

void FieldList():
{}
{
	Field() (LOOKAHEAD(2) FieldSep() Field())* (FieldSep())?
}

void Field():
{}
{
	"[" Exp() "]" "=" Exp() 
|	LOOKAHEAD(2) <NAME> "=" Exp() 
|	Exp()
}

void FieldSep():
{}
{
	"," | ";"
}

void Binop():
{}
{
   "+" | "-" | "*" | "/" | "^" | "%" | ".." | "<" | "<=" | ">" | ">=" | "==" | "~=" | <AND> | <OR>
}

void Unop():
{}
{
	"-" | <NOT> | "#"
}