GhidRust/src/main/java/ghidrust/decompiler/parser/c/gen/c.jjt

/*
C grammar and JJTree definition for use with JavaCC
Contributed by Doug South (dsouth@squirrel.com.au) 21/3/97
From: https://github.com/javacc/javacc/blob/master/grammars/CParser.jj
*/

/* The following code has been significantly modified so as to make it compatible with Ghidra's decompiled output
* and for the Rust transpilation
*/

options {
	MULTI = true;
	VISITOR = true;
	STATIC = false;
}

PARSER_BEGIN(CParser)

import java.io.InputStream;
import java.io.ByteArrayInputStream;
import java.nio.charset.StandardCharsets;
import ghidrust.decompiler.parser.c.CVisitor;

public class CParser {
	private static CParser c_parser;
	
	// Run the parser
	public static String transpile(String c_code) {
		InputStream stream = new ByteArrayInputStream(c_code.getBytes(StandardCharsets.UTF_8));
		return transpile(stream);
	}

	public static String transpile(InputStream stream) {
		c_parser = new CParser(stream);
		
		try {
			return parse();
		} catch (ParseException e) {
			System.out.println("Rust transpiler: Encountered errors during parsing.");
			e.printStackTrace();
			return null;
		}
	}
	
	public static String parse() throws ParseException {
		CParserVisitor visitor = new CVisitor();
        return (String) c_parser.FunctionDefinition().jjtAccept(visitor, null);
	}
}

PARSER_END(CParser)

SKIP : {
	" "
	|  "\t"
	|  "\n"
	|  "\r"
	|  <"//" (~["\n","\r"])* ("\n" | "\r" | "\r\n")>
	|  <"/*" (~["*"])* "*" ("*" | ~["*","/"] (~["*"])* "*")* "/">
	| "#" : PREPROCESSOR_OUTPUT
}

<PREPROCESSOR_OUTPUT> SKIP:
{
	"\n" : DEFAULT
}

<PREPROCESSOR_OUTPUT> MORE:
{
	"\\\n"
	|
	"\\\r\n"
	|
	< ~[] >
}


TOKEN : {
	<INTEGER_LITERAL: <DECIMAL_LITERAL> (["l","L"])? | <HEX_LITERAL> (["l","L"])? | <OCTAL_LITERAL> (["l","L"])?>
	|  <#DECIMAL_LITERAL: ["1"-"9"] (["0"-"9"])*>
	|  <#HEX_LITERAL: "0" ["x","X"] (["0"-"9","a"-"f","A"-"F"])+>
	|  <#OCTAL_LITERAL: "0" (["0"-"7"])*>
	|  <FLOATING_POINT_LITERAL: (["0"-"9"])+ "." (["0"-"9"])* (<EXPONENT>)? (["f","F","d","D"])? | "." (["0"-"9"])+ (<EXPONENT>)? (["f","F","d","D"])? | (["0"-"9"])+ <EXPONENT> (["f","F","d","D"])? | (["0"-"9"])+ (<EXPONENT>)? ["f","F","d","D"]>
	|  <#EXPONENT: ["e","E"] (["+","-"])? (["0"-"9"])+>
	|  <CHARACTER_LITERAL: "\'" (~["\'","\\","\n","\r"] | "\\" (["n","t","b","r","f","\\","\'","\""] | ["0"-"7"] (["0"-"7"])? | ["0"-"3"] ["0"-"7"] ["0"-"7"])) "\'">
	|  <STRING_LITERAL: "\"" ( ~["\"","\\","\n","\r"] | "\\" ( ["n","t","b","r","f","\\","\'","\""] | ["0"-"7"] (["0"-"7"])?  | ["0"-"3"] ["0"-"7"] ["0"-"7"] | ( ["\n","\r"] | "\r\n")))* "\"">
	|  <BOOLEAN: "true" | "false">
	|  <UNDEFINED_TYPE: "undefined" (<DECIMAL_LITERAL>)?> // Ghidra specific unknown type
}

TOKEN : {
	<CONTINUE: "continue"> |
	<VOLATILE: "volatile"> |
	<REGISTER: "register"> |
	<UNSIGNED: "unsigned"> |
	<TYPEDEF: "typedef"> |
	<DFLT: "default"> |
	<DOUBLE: "double"> |
	<SIZEOF: "sizeof"> |
	<SWITCH: "switch"> |
	<RETURN: "return"> |
	<EXTERN: "extern"> |
	<STRUCT: "struct"> |
	<STATIC: "static"> |
	<SIGNED: "signed"> |
	<WHILE: "while"> |
	<BREAK: "break"> |
	<UNION: "union"> |
	<CONST: "const"> |
	<FLOAT: "float"> |
	<SHORT: "short"> |
	<ELSE: "else"> |
	<CASE: "case"> |
	<LONG: "long"> |
	<ENUM: "enum"> |
	<AUTO: "auto"> |
	<VOID: "void"> |
	<CHAR: "char"> |
	<GOTO: "goto"> |
	<FOR: "for"> |
	<INT: "int"> |
	<IF: "if"> |
	<DO: "do"> |
	<BOOL_TYPE: "bool"> |
	<CODE: "code"> // code ptr, Ghidra specific
}

TOKEN : {
	<IDENTIFIER: (<LETTER> | <SEPARATOR> <SEPARATOR>) (<LETTER> | <SEPARATOR> <SEPARATOR> | <DIGIT>)*>
	|  <#LETTER: ["$","A"-"Z","_","a"-"z"]>
	|  <#DIGIT: ["0"-"9"]>
	|  <#SEPARATOR: [":"]>
}


SimpleNode FunctionDefinition() : {}
{
	[LOOKAHEAD(DeclarationSpecifiers()) DeclarationSpecifiers()] Declarator()
	CompoundStatement()
	{
		return jjtThis;
	}
}

void Declaration() : {}
{
	DeclarationSpecifiers() [ InitDeclaratorList() ] ";"
}

void DeclarationList() : {}
{
	( LOOKAHEAD(Declaration()) Declaration() )+
}

void DeclarationSpecifiers() : {}
{
	StorageClassSpecifier() [ LOOKAHEAD(DeclarationSpecifiers())
	DeclarationSpecifiers() ] |
	TypeSpecifier()  [ LOOKAHEAD(DeclarationSpecifiers())
	DeclarationSpecifiers() ] |
	TypeQualifier() [ LOOKAHEAD(DeclarationSpecifiers())
	DeclarationSpecifiers() ]
}

void StorageClassSpecifier() #GhostStringToken :
{
	Token t;
}
{
	( t = <AUTO> | t = <REGISTER> | t = <STATIC> | t = <EXTERN> | t = <TYPEDEF> )
	{
		jjtThis.setValue(t.image);
	}
}

void TypeSpecifier() #TypeStringToken :
{
	Token t;
}
{
	( t = <VOID> | t = <CHAR> | t = <SHORT> | t = <INT> | t = <LONG> | t = <FLOAT> | t = <DOUBLE> | t = <SIGNED> |
	t = <UNSIGNED> | t = <BOOL_TYPE> | t = <CODE> | t = <UNDEFINED_TYPE> )
	{
		jjtThis.setValue(t.image);
	}
}

void TypeQualifier() #StringToken :
{
	Token t;
}
{
	( t = <CONST> | t = <VOLATILE> )
	{
		jjtThis.setValue(t.image);
	}
}

void InitDeclaratorList() : {}
{
	InitDeclarator() ("," InitDeclarator())*
}

void InitDeclarator() : {}
{
	Declarator() [ "=" Initializer() ]
}

void SpecifierQualifierList() : {}
{
	TypeSpecifier() [ LOOKAHEAD(SpecifierQualifierList())
	SpecifierQualifierList() ]|
	TypeQualifier() [ LOOKAHEAD(SpecifierQualifierList())
	SpecifierQualifierList() ]
}

void Declarator() : {}
{
	[ Pointer() ] DirectDeclarator()
}

void DirectDeclarator() : {}
{
	( Identifier() | "(" Declarator() ")" )
	( "[" [ ConstantExpression() ] "]" |
	LOOKAHEAD(3) "(" ParameterTypeList() ")" |
	"(" [ IdentifierList() ] ")" )*
}

void Pointer() : {}
{
	"*" [ TypeQualifierList() ] [ Pointer() ]
}

void TypeQualifierList() : {}
{
	(TypeQualifier())+
}

void ParameterTypeList() : {}
{
	ParameterList() ["," "..." ]
}

void ParameterList() : {}
{
	ParameterDeclaration() (LOOKAHEAD(2) "," ParameterDeclaration())*
}

void ParameterDeclaration() : {}
{
	DeclarationSpecifiers() ( LOOKAHEAD(Declarator()) Declarator() | [ AbstractDeclarator() ] )
}

void IdentifierList() : {}
{
	Identifier() ("," Identifier())*
}

void Initializer() : {}
{
	( AssignmentExpression() |
	"{" InitializerList() [","] "}" )
}

void InitializerList() : {}
{
	Initializer() (LOOKAHEAD(2) "," Initializer())*
}

void TypeName() : {}
{
	SpecifierQualifierList() [ AbstractDeclarator() ]
	
}

void AbstractDeclarator() : {}
{
	( LOOKAHEAD(3) Pointer() |
	[Pointer()] DirectAbstractDeclarator() )
}

void DirectAbstractDeclarator() : {}
{
	( LOOKAHEAD(2) "(" AbstractDeclarator() ")" |
	"[" [ConstantExpression()] "]" |
	"(" [ParameterTypeList()] ")" )
	
	( "[" [ ConstantExpression() ] "]" | "(" [ ParameterTypeList() ] ")" )*
}

void Statement() : {}
{
	( LOOKAHEAD(Identifier() ":") LabeledStatement() |
	ExpressionStatement() |
	CompoundStatement() |
	SelectionStatement() |
	IterationStatement() |
	JumpStatement() )
}

void LabeledStatement() : {}
{
	( Identifier() ":" Statement() |
	<CASE> ConstantExpression() ":" Statement() |
	<DFLT> ":" Statement() )
}

void ExpressionStatement() : {}
{
	[ Expression() ] ";"
}

void CompoundStatement() : {}
{
	"{" [ LOOKAHEAD(DeclarationList()) DeclarationList() ]
	[ StatementList() ]
	"}"
}

void StatementList() : {}
{
	(Statement())+
}

void SelectionStatement() : {}
{
	( <IF> "(" Expression() ")" Statement() [ LOOKAHEAD(2) <ELSE> Statement() ] |
	<SWITCH> "(" Expression() ")" Statement() )
}

void IterationStatement() : {}
{
	( <WHILE> "(" Expression() ")" Statement() |
	<DO> Statement() <WHILE> "(" Expression() ")" ";" |
	<FOR> "(" [ Expression() ] ";" [ Expression() ] ";" [ Expression() ] ")" Statement() )
}

void JumpStatement() : {}
{
	( <GOTO> Identifier() ";" |
	<CONTINUE> ";" |
	<BREAK> ";" |
	<RETURN> [ Expression() ] ";" )
}

void Expression() : {}
{
	AssignmentExpression() ( "," AssignmentExpression() )* | DeclarationSpecifiers() InitDeclaratorList()
}

void AssignmentExpression() : {}
{
	LOOKAHEAD(UnaryExpression() AssignmentOperator()) UnaryExpression() AssignmentOperator() AssignmentExpression() |
	LOOKAHEAD(3) ConditionalExpression()
}

void AssignmentOperator() #StringToken :
{
	Token t;
}
{
	( t = "=" | t = "*=" | t = "/=" | t = "%=" | t = "+=" | t = "-=" | t = "<<=" | t = ">>=" | t = "&=" | t = "^=" | t = "|=" )
	{
		jjtThis.setValue(t.image);
	}
}

void ConditionalExpression() : {}
{
	LogicalORExpression() [ "?" Expression() ":" ConditionalExpression() ]
}

void ConstantExpression() : {}
{
	ConditionalExpression()
}

void LogicalORExpression() : {}
{
	LogicalANDExpression() [ "||" LogicalORExpression() ]
}

void LogicalANDExpression() : {}
{
	InclusiveORExpression() [ "&&" LogicalANDExpression() ]
}

void InclusiveORExpression() : {}
{
	ExclusiveORExpression() [ "|" InclusiveORExpression() ]
}

void ExclusiveORExpression() : {}
{
	ANDExpression() [ "^" ExclusiveORExpression() ]
}

void ANDExpression() : {}
{
	EqualityExpression() [ "&" ANDExpression() ]
}

void EqualityExpression() : {}
{
	RelationalExpression() [ ( "==" | "!=" ) EqualityExpression() ]
}

void RelationalExpression() : {}
{
	ShiftExpression() [ ( "<" | ">" | "<=" | ">=" ) RelationalExpression() ]
}

void ShiftExpression() : {}
{
	AdditiveExpression() [ ( "<<" | ">>" ) ShiftExpression() ]
}

void AdditiveExpression() : {}
{
	MultiplicativeExpression() [ AdditionOperator() AdditiveExpression() ]
}

void AdditionOperator() #StringToken :
{
	Token t;
}
{
	( t = "+" | t = "-" )
	{
		jjtThis.setValue(t.image);
	}
}

void MultiplicativeExpression() : {}
{
	CastExpression() [ ( "*" | "/" | "%" ) MultiplicativeExpression() ]
}

void CastExpression() : {}
{
	( LOOKAHEAD("(" TypeName() ")" CastExpression() ) "(" TypeName() ")" CastExpression() |
	UnaryExpression() )
}

void UnaryExpression() : {}
{
	( LOOKAHEAD(3) PostfixExpression() |
	"++" UnaryExpression() |
	"--" UnaryExpression() |
	UnaryOperator() CastExpression() |
	<SIZEOF> ( LOOKAHEAD(UnaryExpression() ) UnaryExpression() | "(" TypeName() ")" ) )
}

void UnaryOperator() : {}
{
	( "&" | "*" | "+" | "-" | "~" | "!" )
}

void PostfixExpression() : {
	int choice = 0;
}
{
	PrimaryExpression() ( "[" Expression() "]" { choice = 1; } |
	"(" [ LOOKAHEAD(ArgumentExpressionList() ) ArgumentExpressionList() ] ")" { choice = 2; } |
	"." Identifier() { choice = 3; } |
	"->" Identifier() | { choice = 4; }
	"++" | { choice = 5; }
	"--" { choice = 6; } )*
	{
		jjtThis.choice = choice;
	}
}

void PrimaryExpression() : {}
{
	( Identifier() |
	Constant() |
	"(" Expression() ")" )
}

void ArgumentExpressionList() : {}
{
	AssignmentExpression() ( "," AssignmentExpression() )*
}

ASTStringToken Identifier() #StringToken :
{
	Token r = null;
	ASTStringToken t = null;
	Token s = null;
	int choice;
}
{
	(LOOKAHEAD(<IDENTIFIER>) r = <IDENTIFIER> { choice = 0; } | [ r = <IDENTIFIER> ] "<" t = Identifier() ">" [ s = <IDENTIFIER> ] { choice = 1; })
	{
		if (choice == 1) {
			jjtThis.setValue((r != null ? r.image : "") + "<" + t.getValue() + ">" + (s != null ? s.image : ""));
		} else {
			jjtThis.setValue(r.image);
		}

		return jjtThis;
	}
}

void Constant() #StringToken :
{
	Token t;
}
{
	(t = <INTEGER_LITERAL> | t = <FLOATING_POINT_LITERAL> | t = <CHARACTER_LITERAL> | t = <STRING_LITERAL> | t = <BOOLEAN>)
	{
		jjtThis.setValue(t.image);
	}
}