diff --git a/ast/ast.go b/ast/ast.go index 34540b1..fbbbf70 100644 --- a/ast/ast.go +++ b/ast/ast.go @@ -1,9 +1,13 @@ package ast -import "monkey/token" +import ( + "bytes" + "monkey/token" +) type Node interface { TokenLiteral() string + String() string } type Statement interface { @@ -28,6 +32,16 @@ func (p *Program) TokenLiteral() string { } } +func (p *Program) String() string { + var out bytes.Buffer + + for _, s := range p.Statements { + out.WriteString(s.String()) + } + + return out.String() +} + type Identifier struct { Token token.Token // the token.Ident token Value string @@ -44,20 +58,102 @@ type ReturnStatement struct { ReturnValue Expression } +type ExpressionStatement struct { + Token token.Token // the first token of the expression + Expression Expression +} + +type IntegerLiteral struct { + Token token.Token + Value int64 +} + +type PrefixExpression struct { + Token token.Token + Operator string + Right Expression +} + func (ls *LetStatement) statementNode() { } func (ls *LetStatement) TokenLiteral() string { return ls.Token.Literal } +func (ls *LetStatement) String() string { + var out bytes.Buffer + + out.WriteString(ls.TokenLiteral() + " ") + out.WriteString(ls.Name.String()) + out.WriteString(" = ") + + if ls.Value != nil { + out.WriteString(ls.Value.String()) + } + + out.WriteString(";") + + return out.String() +} func (i *Identifier) expressionNode() { } func (i *Identifier) TokenLiteral() string { return i.Token.Literal } +func (i *Identifier) String() string { return i.Value } -func (i *ReturnStatement) statementNode() { +func (rs *ReturnStatement) statementNode() { } -func (i *ReturnStatement) TokenLiteral() string { - return i.Token.Literal +func (rs *ReturnStatement) TokenLiteral() string { + return rs.Token.Literal +} +func (rs *ReturnStatement) String() string { + var out bytes.Buffer + + out.WriteString(rs.TokenLiteral() + " ") + + if rs.ReturnValue != nil { + out.WriteString(rs.ReturnValue.String()) + } + + out.WriteString(";") + + return out.String() +} + +func (es *ExpressionStatement) statementNode() { +} +func (es *ExpressionStatement) TokenLiteral() string { + return es.Token.Literal +} +func (es *ExpressionStatement) String() string { + if es.Expression != nil { + return es.Expression.String() + } + + return "" +} + +func (il *IntegerLiteral) expressionNode() { +} +func (il *IntegerLiteral) TokenLiteral() string { + return il.Token.Literal +} +func (il *IntegerLiteral) String() string { + return il.Token.Literal +} + +func (pe *PrefixExpression) expressionNode() {} +func (pe *PrefixExpression) TokenLiteral() string { + return pe.Token.Literal +} +func (pe *PrefixExpression) String() string { + var out bytes.Buffer + + out.WriteString("(") + out.WriteString(pe.Operator) + out.WriteString(pe.Right.String()) + out.WriteString(")") + + return out.String() } diff --git a/ast/ast_test.go b/ast/ast_test.go new file mode 100644 index 0000000..e1bea83 --- /dev/null +++ b/ast/ast_test.go @@ -0,0 +1,28 @@ +package ast + +import ( + "monkey/token" + "testing" +) + +func TestString(t *testing.T) { + program := &Program{ + Statements: []Statement{ + &LetStatement{ + Token: token.Token{Type: token.LET, Literal: "let"}, + Name: &Identifier{ + Token: token.Token{Type: token.IDENT, Literal: "myVar"}, + Value: "myVar", + }, + Value: &Identifier{ + Token: token.Token{Type: token.IDENT, Literal: "anotherVar"}, + Value: "anotherVar", + }, + }, + }, + } + + if program.String() != "let myVar = anotherVar;" { + t.Errorf("program.String wrong. got=%q", program.String()) + } +} diff --git a/parser/parser.go b/parser/parser.go index ee8e718..a42dd6a 100644 --- a/parser/parser.go +++ b/parser/parser.go @@ -5,6 +5,18 @@ import ( "monkey/ast" "monkey/lexer" "monkey/token" + "strconv" +) + +const ( + _ int = iota + LOWEST + EQUALS // == + LESSGREATER // > or < + SUM // + + PRODUCT // * + PREFIX // -X or !X + CALL // myFunction(X) ) type Parser struct { @@ -13,8 +25,16 @@ type Parser struct { curToken token.Token peekToken token.Token + + prefixParseFns map[token.TokenType]prefixParseFn + infixParseFns map[token.TokenType]infixParseFn } +type ( + prefixParseFn func() ast.Expression + infixParseFn func(expression ast.Expression) ast.Expression +) + func (p *Parser) nextToken() { p.curToken = p.peekToken p.peekToken = p.l.NextToken() @@ -26,6 +46,12 @@ func New(l *lexer.Lexer) *Parser { errors: []string{}, } + p.prefixParseFns = make(map[token.TokenType]prefixParseFn) + p.registerPrefix(token.IDENT, p.parseIdentifier) + p.registerPrefix(token.INT, p.parseIntegerLiteral) + p.registerPrefix(token.BANG, p.parsePrefixExpression) + p.registerPrefix(token.MINUS, p.parsePrefixExpression) + // Read two token, so curToken and peekToken are both set p.nextToken() p.nextToken() @@ -55,7 +81,7 @@ func (p *Parser) parseStatement() ast.Statement { case token.RETURN: return p.parseReturnStatement() default: - return nil + return p.parseExpressionsStatement() } } @@ -123,3 +149,74 @@ func (p *Parser) parseReturnStatement() *ast.ReturnStatement { return stmt } + +func (p *Parser) parseExpressionsStatement() *ast.ExpressionStatement { + stmt := &ast.ExpressionStatement{Token: p.curToken} + + stmt.Expression = p.parseExpression(LOWEST) + + if p.peekTokenIs(token.SEMICOLON) { + p.nextToken() + } + + return stmt +} + +func (p *Parser) parseExpression(precedence int) ast.Expression { + prefix := p.prefixParseFns[p.curToken.Type] + if prefix == nil { + p.noPrefixParseFnError(p.curToken.Type) + return nil + } + leftExp := prefix() + + return leftExp +} + +func (p *Parser) parseIntegerLiteral() ast.Expression { + lit := &ast.IntegerLiteral{Token: p.curToken} + + value, err := strconv.ParseInt(p.curToken.Literal, 0, 64) + if err != nil { + msg := fmt.Sprintf("could not parse %q as integer", p.curToken.Literal) + p.errors = append(p.errors, msg) + return nil + } + + lit.Value = value + + return lit +} + +func (p *Parser) registerPrefix(tokenType token.TokenType, fn prefixParseFn) { + p.prefixParseFns[tokenType] = fn +} + +func (p *Parser) registerInfix(tokenType token.TokenType, fn infixParseFn) { + p.infixParseFns[tokenType] = fn +} + +func (p *Parser) parseIdentifier() ast.Expression { + return &ast.Identifier{ + Token: p.curToken, + Value: p.curToken.Literal, + } +} + +func (p *Parser) noPrefixParseFnError(t token.TokenType) { + msg := fmt.Sprintf("no prefix parse function for %s found", t) + p.errors = append(p.errors, msg) +} + +func (p *Parser) parsePrefixExpression() ast.Expression { + expression := &ast.PrefixExpression{ + Token: p.curToken, + Operator: p.curToken.Literal, + } + + p.nextToken() + + expression.Right = p.parseExpression(PREFIX) + + return expression +} diff --git a/parser/parser_test.go b/parser/parser_test.go index f6c317f..69023ad 100644 --- a/parser/parser_test.go +++ b/parser/parser_test.go @@ -1,6 +1,7 @@ package parser import ( + "fmt" "monkey/ast" "monkey/lexer" "testing" @@ -73,3 +74,116 @@ func testLetStatements(t *testing.T, s ast.Statement, name string) bool { return true } + +func TestIdentifierExpressions(t *testing.T) { + input := "foobar;" + + l := lexer.New(input) + p := New(l) + program := p.ParseProgram() + checkParserErrors(t, p) + + if len(program.Statements) != 1 { + t.Fatalf("program has not enough statements. got=%d", len(program.Statements)) + } + stmt, ok := program.Statements[0].(*ast.ExpressionStatement) + if !ok { + t.Fatalf("program.Statements[0] is not ast.ExpressionStatement. got=%T", program.Statements[0]) + } + + ident, ok := stmt.Expression.(*ast.Identifier) + if !ok { + t.Fatalf("exp is not *ast.Identifier. got=%T", stmt.Expression) + } + if ident.Value != "foobar" { + t.Errorf("ident.Value not %s. got=%s", "foobar", ident.Value) + } + if ident.TokenLiteral() != "foobar" { + t.Errorf("ident.TokenLiteral not %s. got=%s", "foobar", ident.TokenLiteral()) + } +} + +func TestIntegerLiteralExpressions(t *testing.T) { + input := "5;" + + l := lexer.New(input) + p := New(l) + program := p.ParseProgram() + checkParserErrors(t, p) + + if len(program.Statements) != 1 { + t.Fatalf("program has not enough statements. got=%d", len(program.Statements)) + } + stmt, ok := program.Statements[0].(*ast.ExpressionStatement) + if !ok { + t.Fatalf("program.Statements[0] is not ast.ExpressionStatement. got=%T", program.Statements[0]) + } + + literal, ok := stmt.Expression.(*ast.IntegerLiteral) + if !ok { + t.Fatalf("exp is not *ast.IntegerLiteral. got=%T", stmt.Expression) + } + if literal.Value != 5 { + t.Errorf("ident.Value not %d. got=%d", 5, literal.Value) + } + if literal.TokenLiteral() != "5" { + t.Errorf("ident.TokenLiteral not %s. got=%s", "5", literal.TokenLiteral()) + } +} + +func TestParsingPrefixExpressions(t *testing.T) { + prefixTests := []struct { + input string + operator string + integerValue int64 + }{ + {"!5;", "!", 5}, + {"-15;", "-", 15}, + } + + for _, tt := range prefixTests { + l := lexer.New(tt.input) + p := New(l) + program := p.ParseProgram() + checkParserErrors(t, p) + + if len(program.Statements) != 1 { + t.Fatalf("program.Statements does not contain %d statements. got=%d\n", 1, len(program.Statements)) + } + stmt, ok := program.Statements[0].(*ast.ExpressionStatement) + if !ok { + t.Fatalf("program.Statements[0] is not ast.ExpressionStatement. got=%T", program.Statements[0]) + } + + exp, ok := stmt.Expression.(*ast.PrefixExpression) + if !ok { + t.Fatalf("exp is not *ast.PrefixExpression. got=%T", stmt.Expression) + } + if exp.Operator != tt.operator { + t.Fatalf("exp.Operator is not '%s'. got=%s", tt.operator, exp.Operator) + } + if !testIntegerLiteral(t, exp.Right, tt.integerValue) { + return + } + } +} + +func testIntegerLiteral(t *testing.T, il ast.Expression, value int64) bool { + integ, ok := il.(*ast.IntegerLiteral) + if !ok { + t.Errorf("il not *ast.IntegerLiteral. got=%T", il) + return false + } + + if integ.Value != value { + t.Errorf("integ.Value not %d. got=%d", value, integ.Value) + return false + } + + if integ.TokenLiteral() != fmt.Sprintf("%d", value) { + t.Errorf("integ.TopkenLiteral not %d. got=%s", value, integ.TokenLiteral()) + return false + } + + return true +}