Sharkk/Mako
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Compare commits

base: Sharkk:121774005b6aea6e877e38802bafb3686a15b9c8
Branches Tags
Sharkk:master
...
compare: Sharkk:6575b077fee557b8d1acda4181646e5c2dbbce99
Branches Tags
Sharkk:master

2 Commits
121774005b ... 6575b077fe

Author SHA1 Message Date
Sky Johnson
6575b077fe symbol pre-resolution 2025-06-11 23:19:54 -05:00
Sky Johnson
3d07d010dc modulo operator 2025-06-11 23:08:34 -05:00
6 changed files with 396 additions and 168 deletions
Show Stats Expand all files Collapse all files

217
parser/ast.go
View File

@ -2,11 +2,16 @@ package parser
import "fmt"
// Note: Type definitions moved to types.go for proper separation of concerns
// Position represents source location information
type Position struct {
Line int
Column int
}
// Node represents any node in the AST
type Node interface {
String() string
Pos() Position
}
// Statement represents statement nodes that can appear at the top level or in blocks
@ -19,14 +24,14 @@ type Statement interface {
type Expression interface {
Node
expressionNode()
TypeInfo() TypeInfo // Returns type by value, not pointer
TypeInfo() TypeInfo
}
// Program represents the root of the AST containing all top-level statements.
// Tracks exit code for script termination and owns the statement list.
type Program struct {
Statements []Statement
ExitCode int
Position Position
}
func (p *Program) String() string {
@ -36,24 +41,26 @@ func (p *Program) String() string {
}
return result
}
func (p *Program) Pos() Position { return p.Position }
// StructField represents a field definition within a struct.
// Contains field name and required type annotation for compile-time checking.
type StructField struct {
Name string
TypeHint TypeInfo // Required for struct fields, embeds directly
TypeHint TypeInfo
Position Position
}
func (sf *StructField) String() string {
return fmt.Sprintf("%s: %s", sf.Name, typeToString(sf.TypeHint))
}
func (sf *StructField) Pos() Position { return sf.Position }
// StructStatement represents struct type definitions with named fields.
// Defines new types that can be instantiated and used for type checking.
type StructStatement struct {
Name string
Fields []StructField
ID uint16 // Unique identifier for fast lookup
ID uint16
Position Position
}
func (ss *StructStatement) statementNode() {}
@ -67,26 +74,28 @@ func (ss *StructStatement) String() string {
}
return fmt.Sprintf("struct %s {\n\t%s\n}", ss.Name, fields)
}
func (ss *StructStatement) Pos() Position { return ss.Position }
// MethodDefinition represents method definitions attached to struct types.
// Links a function implementation to a specific struct via struct ID.
type MethodDefinition struct {
StructID uint16 // Index into struct table for fast lookup
StructID uint16
MethodName string
Function *FunctionLiteral
Position Position
}
func (md *MethodDefinition) statementNode() {}
func (md *MethodDefinition) String() string {
return fmt.Sprintf("fn <struct>.%s%s", md.MethodName, md.Function.String()[2:])
}
func (md *MethodDefinition) Pos() Position { return md.Position }
// StructConstructor represents struct instantiation with field initialization.
// Uses struct ID for fast type resolution and validation during parsing.
type StructConstructor struct {
StructID uint16 // Index into struct table
Fields []TablePair // Reuses table pair structure for field assignments
typeInfo TypeInfo // Cached type info for this constructor
StructID uint16
Fields []TablePair
typeInfo TypeInfo
Position Position
}
func (sc *StructConstructor) expressionNode() {}
@ -98,15 +107,16 @@ func (sc *StructConstructor) String() string {
return fmt.Sprintf("<struct>{%s}", joinStrings(pairs, ", "))
}
func (sc *StructConstructor) TypeInfo() TypeInfo { return sc.typeInfo }
func (sc *StructConstructor) Pos() Position { return sc.Position }
// Assignment represents both variable assignment statements and assignment expressions.
// Unified design reduces AST node count and simplifies type checking logic.
type Assignment struct {
Target Expression // Target (identifier, dot, or index expression)
Value Expression // Value being assigned
TypeHint TypeInfo // Optional explicit type hint, embeds directly
IsDeclaration bool // True if declaring new variable in current scope
IsExpression bool // True if used as expression (wrapped in parentheses)
Target Expression
Value Expression
TypeHint TypeInfo
IsDeclaration bool
IsExpression bool
Position Position
}
func (a *Assignment) statementNode() {}
@ -131,40 +141,45 @@ func (a *Assignment) String() string {
return result
}
func (a *Assignment) TypeInfo() TypeInfo { return a.Value.TypeInfo() }
func (a *Assignment) Pos() Position { return a.Position }
// ExpressionStatement wraps expressions used as statements.
// Allows function calls and other expressions at statement level.
type ExpressionStatement struct {
Expression Expression
Position Position
}
func (es *ExpressionStatement) statementNode() {}
func (es *ExpressionStatement) String() string {
return es.Expression.String()
}
func (es *ExpressionStatement) Pos() Position { return es.Position }
// EchoStatement represents output statements for displaying values.
// Simple debugging and output mechanism built into the language.
type EchoStatement struct {
Value Expression
Position Position
}
func (es *EchoStatement) statementNode() {}
func (es *EchoStatement) String() string {
return fmt.Sprintf("echo %s", es.Value.String())
}
func (es *EchoStatement) Pos() Position { return es.Position }
// BreakStatement represents loop exit statements.
// Simple marker node with no additional data needed.
type BreakStatement struct{}
type BreakStatement struct {
Position Position
}
func (bs *BreakStatement) statementNode() {}
func (bs *BreakStatement) String() string { return "break" }
func (bs *BreakStatement) Pos() Position { return bs.Position }
// ExitStatement represents script termination with optional exit code.
// Value expression is nil for plain "exit", non-nil for "exit <code>".
type ExitStatement struct {
Value Expression // Optional exit code expression
Value Expression
Position Position
}
func (es *ExitStatement) statementNode() {}
@ -174,11 +189,12 @@ func (es *ExitStatement) String() string {
}
return fmt.Sprintf("exit %s", es.Value.String())
}
func (es *ExitStatement) Pos() Position { return es.Position }
// ReturnStatement represents function return with optional value.
// Value expression is nil for plain "return", non-nil for "return <value>".
type ReturnStatement struct {
Value Expression // Optional return value expression
Value Expression
Position Position
}
func (rs *ReturnStatement) statementNode() {}
@ -188,12 +204,13 @@ func (rs *ReturnStatement) String() string {
}
return fmt.Sprintf("return %s", rs.Value.String())
}
func (rs *ReturnStatement) Pos() Position { return rs.Position }
// ElseIfClause represents conditional branches in if statements.
// Contains condition expression and body statements for this branch.
type ElseIfClause struct {
Condition Expression
Body []Statement
Position Position
}
func (eic *ElseIfClause) String() string {
@ -203,14 +220,15 @@ func (eic *ElseIfClause) String() string {
}
return fmt.Sprintf("elseif %s then\n%s", eic.Condition.String(), body)
}
func (eic *ElseIfClause) Pos() Position { return eic.Position }
// IfStatement represents conditional execution with optional elseif and else branches.
// Supports multiple elseif clauses and an optional final else clause.
type IfStatement struct {
Condition Expression // Main condition
Body []Statement // Statements to execute if condition is true
ElseIfs []ElseIfClause // Optional elseif branches
Else []Statement // Optional else branch
Condition Expression
Body []Statement
ElseIfs []ElseIfClause
Else []Statement
Position Position
}
func (is *IfStatement) statementNode() {}
@ -236,12 +254,13 @@ func (is *IfStatement) String() string {
result += "end"
return result
}
func (is *IfStatement) Pos() Position { return is.Position }
// WhileStatement represents condition-based loops that execute while condition is true.
// Contains condition expression and body statements to repeat.
type WhileStatement struct {
Condition Expression
Body []Statement
Position Position
}
func (ws *WhileStatement) statementNode() {}
@ -256,15 +275,16 @@ func (ws *WhileStatement) String() string {
result += "end"
return result
}
func (ws *WhileStatement) Pos() Position { return ws.Position }
// ForStatement represents numeric for loops with start, end, and optional step.
// Variable is automatically scoped to the loop body.
type ForStatement struct {
Variable *Identifier // Loop variable (automatically number type)
Start Expression // Starting value expression
End Expression // Ending value expression
Step Expression // Optional step expression (nil means step of 1)
Body []Statement // Loop body statements
Variable *Identifier
Start Expression
End Expression
Step Expression
Body []Statement
Position Position
}
func (fs *ForStatement) statementNode() {}
@ -285,14 +305,15 @@ func (fs *ForStatement) String() string {
result += "end"
return result
}
func (fs *ForStatement) Pos() Position { return fs.Position }
// ForInStatement represents iterator-based loops over tables, arrays, or other iterables.
// Supports both single variable (for v in iter) and key-value (for k,v in iter) forms.
type ForInStatement struct {
Key *Identifier // Optional key variable (nil for single variable iteration)
Value *Identifier // Value variable (required)
Iterable Expression // Expression to iterate over
Body []Statement // Loop body statements
Key *Identifier
Value *Identifier
Iterable Expression
Body []Statement
Position Position
}
func (fis *ForInStatement) statementNode() {}
@ -313,12 +334,13 @@ func (fis *ForInStatement) String() string {
result += "end"
return result
}
func (fis *ForInStatement) Pos() Position { return fis.Position }
// FunctionParameter represents a parameter in function definitions.
// Contains parameter name and optional type hint for type checking.
type FunctionParameter struct {
Name string
TypeHint TypeInfo // Optional type constraint, embeds directly
TypeHint TypeInfo
Position Position
}
func (fp *FunctionParameter) String() string {
@ -327,12 +349,15 @@ func (fp *FunctionParameter) String() string {
}
return fp.Name
}
func (fp *FunctionParameter) Pos() Position { return fp.Position }
// Identifier represents variable references and names.
// Stores resolved type information for efficient type checking.
type Identifier struct {
Value string
typeInfo TypeInfo // Resolved type, embeds directly
ScopeDepth int // 0 = global, 1+ = local depth
SlotIndex int // register/stack slot (-1 = unresolved)
typeInfo TypeInfo
Position Position
}
func (i *Identifier) expressionNode() {}
@ -343,31 +368,35 @@ func (i *Identifier) TypeInfo() TypeInfo {
}
return i.typeInfo
}
func (i *Identifier) Pos() Position { return i.Position }
func (i *Identifier) IsResolved() bool { return i.SlotIndex >= 0 }
// NumberLiteral represents numeric constants including integers, floats, hex, and binary.
// Always has number type, so no additional type storage needed.
type NumberLiteral struct {
Value float64 // All numbers stored as float64 for simplicity
Value float64
Position Position
}
func (nl *NumberLiteral) expressionNode() {}
func (nl *NumberLiteral) String() string { return fmt.Sprintf("%.2f", nl.Value) }
func (nl *NumberLiteral) TypeInfo() TypeInfo { return NumberType }
func (nl *NumberLiteral) Pos() Position { return nl.Position }
// StringLiteral represents string constants and multiline strings.
// Always has string type, so no additional type storage needed.
type StringLiteral struct {
Value string // String content without quotes
Value string
Position Position
}
func (sl *StringLiteral) expressionNode() {}
func (sl *StringLiteral) String() string { return fmt.Sprintf(`"%s"`, sl.Value) }
func (sl *StringLiteral) TypeInfo() TypeInfo { return StringType }
func (sl *StringLiteral) Pos() Position { return sl.Position }
// BooleanLiteral represents true and false constants.
// Always has bool type, so no additional type storage needed.
type BooleanLiteral struct {
Value bool
Position Position
}
func (bl *BooleanLiteral) expressionNode() {}
@ -378,22 +407,28 @@ func (bl *BooleanLiteral) String() string {
return "false"
}
func (bl *BooleanLiteral) TypeInfo() TypeInfo { return BoolType }
func (bl *BooleanLiteral) Pos() Position { return bl.Position }
// NilLiteral represents the nil constant value.
// Always has nil type, so no additional type storage needed.
type NilLiteral struct{}
type NilLiteral struct {
Position Position
}
func (nl *NilLiteral) expressionNode() {}
func (nl *NilLiteral) String() string { return "nil" }
func (nl *NilLiteral) TypeInfo() TypeInfo { return NilType }
func (nl *NilLiteral) Pos() Position { return nl.Position }
// FunctionLiteral represents function definitions with parameters, body, and optional return type.
// Always has function type, stores additional return type information separately.
type FunctionLiteral struct {
Parameters []FunctionParameter // Function parameters with optional types
Body []Statement // Function body statements
ReturnType TypeInfo // Optional return type hint, embeds directly
Variadic bool // True if function accepts variable arguments
Parameters []FunctionParameter
Body []Statement
ReturnType TypeInfo
Variadic bool
LocalCount int // Pre-computed local variable count
UpvalueCount int // Number of captured variables
MaxStackDepth int // Maximum expression evaluation depth
Position Position
}
func (fl *FunctionLiteral) expressionNode() {}
@ -425,13 +460,14 @@ func (fl *FunctionLiteral) String() string {
return result
}
func (fl *FunctionLiteral) TypeInfo() TypeInfo { return FunctionType }
func (fl *FunctionLiteral) Pos() Position { return fl.Position }
// CallExpression represents function calls with arguments.
// Stores inferred return type from function signature analysis.
type CallExpression struct {
Function Expression // Function expression to call
Arguments []Expression // Argument expressions
typeInfo TypeInfo // Inferred return type, embeds directly
Function Expression
Arguments []Expression
typeInfo TypeInfo
Position Position
}
func (ce *CallExpression) expressionNode() {}
@ -443,13 +479,14 @@ func (ce *CallExpression) String() string {
return fmt.Sprintf("%s(%s)", ce.Function.String(), joinStrings(args, ", "))
}
func (ce *CallExpression) TypeInfo() TypeInfo { return ce.typeInfo }
func (ce *CallExpression) Pos() Position { return ce.Position }
// PrefixExpression represents unary operations like negation and logical not.
// Stores result type based on operator and operand type analysis.
type PrefixExpression struct {
Operator string // Operator symbol ("-", "not")
Right Expression // Operand expression
typeInfo TypeInfo // Result type, embeds directly
Operator string
Right Expression
typeInfo TypeInfo
Position Position
}
func (pe *PrefixExpression) expressionNode() {}
@ -460,14 +497,15 @@ func (pe *PrefixExpression) String() string {
return fmt.Sprintf("(%s%s)", pe.Operator, pe.Right.String())
}
func (pe *PrefixExpression) TypeInfo() TypeInfo { return pe.typeInfo }
func (pe *PrefixExpression) Pos() Position { return pe.Position }
// InfixExpression represents binary operations between two expressions.
// Stores result type based on operator and operand type compatibility.
type InfixExpression struct {
Left Expression // Left operand
Right Expression // Right operand
Operator string // Operator symbol ("+", "-", "==", "and", etc.)
typeInfo TypeInfo // Result type, embeds directly
Left Expression
Right Expression
Operator string
typeInfo TypeInfo
Position Position
}
func (ie *InfixExpression) expressionNode() {}
@ -475,13 +513,14 @@ func (ie *InfixExpression) String() string {
return fmt.Sprintf("(%s %s %s)", ie.Left.String(), ie.Operator, ie.Right.String())
}
func (ie *InfixExpression) TypeInfo() TypeInfo { return ie.typeInfo }
func (ie *InfixExpression) Pos() Position { return ie.Position }
// IndexExpression represents bracket-based member access (table[key]).
// Stores inferred element type based on container type analysis.
type IndexExpression struct {
Left Expression // Container expression
Index Expression // Index/key expression
typeInfo TypeInfo // Element type, embeds directly
Left Expression
Index Expression
typeInfo TypeInfo
Position Position
}
func (ie *IndexExpression) expressionNode() {}
@ -489,13 +528,14 @@ func (ie *IndexExpression) String() string {
return fmt.Sprintf("%s[%s]", ie.Left.String(), ie.Index.String())
}
func (ie *IndexExpression) TypeInfo() TypeInfo { return ie.typeInfo }
func (ie *IndexExpression) Pos() Position { return ie.Position }
// DotExpression represents dot-based member access (table.key).
// Stores inferred member type based on container type and field analysis.
type DotExpression struct {
Left Expression // Container expression
Key string // Member name
typeInfo TypeInfo // Member type, embeds directly
Left Expression
Key string
typeInfo TypeInfo
Position Position
}
func (de *DotExpression) expressionNode() {}
@ -503,12 +543,13 @@ func (de *DotExpression) String() string {
return fmt.Sprintf("%s.%s", de.Left.String(), de.Key)
}
func (de *DotExpression) TypeInfo() TypeInfo { return de.typeInfo }
func (de *DotExpression) Pos() Position { return de.Position }
// TablePair represents key-value pairs in table literals and struct constructors.
// Key is nil for array-style elements, non-nil for object-style elements.
type TablePair struct {
Key Expression // Key expression (nil for array elements)
Value Expression // Value expression
Key Expression
Value Expression
Position Position
}
func (tp *TablePair) String() string {
@ -517,11 +558,12 @@ func (tp *TablePair) String() string {
}
return fmt.Sprintf("%s = %s", tp.Key.String(), tp.Value.String())
}
func (tp *TablePair) Pos() Position { return tp.Position }
// TableLiteral represents table/array/object literals with key-value pairs.
// Always has table type, provides methods to check if it's array-style.
type TableLiteral struct {
Pairs []TablePair // Key-value pairs (key nil for array elements)
Pairs []TablePair
Position Position
}
func (tl *TableLiteral) expressionNode() {}
@ -533,6 +575,7 @@ func (tl *TableLiteral) String() string {
return fmt.Sprintf("{%s}", joinStrings(pairs, ", "))
}
func (tl *TableLiteral) TypeInfo() TypeInfo { return TableType }
func (tl *TableLiteral) Pos() Position { return tl.Position }
// IsArray returns true if this table contains only array-style elements (no explicit keys)
func (tl *TableLiteral) IsArray() bool {

2
parser/lexer.go
View File

@ -257,6 +257,8 @@ func (l *Lexer) NextToken() Token {
tok = Token{Type: STAR, Literal: string(l.ch), Line: l.line, Column: l.column}
case '/':
tok = Token{Type: SLASH, Literal: string(l.ch), Line: l.line, Column: l.column}
case '%':
tok = Token{Type: MOD, Literal: string(l.ch), Line: l.line, Column: l.column}
case ':':
tok = Token{Type: COLON, Literal: string(l.ch), Line: l.line, Column: l.column}
case '.':

261
parser/parser.go
View File

@ -31,9 +31,11 @@ type Parser struct {
errors []ParseError
// Scope tracking
scopes []map[string]bool
// Scope tracking with slot management
scopes []map[string]*Variable
scopeTypes []string
scopeSlots []int // Next available slot per scope
currentDepth int
// Struct tracking with ID mapping
structs map[string]*StructStatement
@ -41,13 +43,23 @@ type Parser struct {
nextID uint16
}
// Variable represents a resolved variable in scope
type Variable struct {
Name string
ScopeDepth int
SlotIndex int
TypeHint TypeInfo
}
// NewParser creates a new parser instance
func NewParser(lexer *Lexer) *Parser {
p := &Parser{
lexer: lexer,
errors: []ParseError{},
scopes: []map[string]bool{make(map[string]bool)},
scopes: []map[string]*Variable{make(map[string]*Variable)},
scopeTypes: []string{"global"},
scopeSlots: []int{0},
currentDepth: 0,
structs: make(map[string]*StructStatement),
structIDs: make(map[uint16]*StructStatement),
nextID: 1, // 0 reserved for non-struct types
@ -71,6 +83,7 @@ func NewParser(lexer *Lexer) *Parser {
p.registerInfix(MINUS, p.parseInfixExpression)
p.registerInfix(SLASH, p.parseInfixExpression)
p.registerInfix(STAR, p.parseInfixExpression)
p.registerInfix(MOD, p.parseInfixExpression)
p.registerInfix(EQ, p.parseInfixExpression)
p.registerInfix(NOT_EQ, p.parseInfixExpression)
p.registerInfix(LT, p.parseInfixExpression)
@ -90,6 +103,11 @@ func NewParser(lexer *Lexer) *Parser {
return p
}
// Helper to create position from token
func (p *Parser) pos() Position {
return Position{Line: p.curToken.Line, Column: p.curToken.Column}
}
// Struct management
func (p *Parser) registerStruct(stmt *StructStatement) {
stmt.ID = p.nextID
@ -107,20 +125,24 @@ func (p *Parser) isStructDefined(name string) bool {
return exists
}
// Scope management
// Scope management with slot allocation
func (p *Parser) enterScope(scopeType string) {
p.scopes = append(p.scopes, make(map[string]bool))
p.scopes = append(p.scopes, make(map[string]*Variable))
p.scopeTypes = append(p.scopeTypes, scopeType)
p.scopeSlots = append(p.scopeSlots, 0)
p.currentDepth++
}
func (p *Parser) exitScope() {
if len(p.scopes) > 1 {
p.scopes = p.scopes[:len(p.scopes)-1]
p.scopeTypes = p.scopeTypes[:len(p.scopeTypes)-1]
p.scopeSlots = p.scopeSlots[:len(p.scopeSlots)-1]
p.currentDepth--
}
}
func (p *Parser) currentVariableScope() map[string]bool {
func (p *Parser) currentVariableScope() map[string]*Variable {
if len(p.scopeTypes) > 1 && p.scopeTypes[len(p.scopeTypes)-1] == "loop" {
return p.scopes[len(p.scopes)-2]
}
@ -129,19 +151,55 @@ func (p *Parser) currentVariableScope() map[string]bool {
func (p *Parser) isVariableDeclared(name string) bool {
for i := len(p.scopes) - 1; i >= 0; i-- {
if p.scopes[i][name] {
if _, exists := p.scopes[i][name]; exists {
return true
}
}
return false
}
func (p *Parser) declareVariable(name string) {
p.currentVariableScope()[name] = true
func (p *Parser) declareVariable(name string, typeHint TypeInfo) *Variable {
scope := p.currentVariableScope()
scopeIdx := len(p.scopes) - 1
if len(p.scopeTypes) > 1 && p.scopeTypes[len(p.scopeTypes)-1] == "loop" {
scopeIdx--
}
variable := &Variable{
Name: name,
ScopeDepth: p.currentDepth,
SlotIndex: p.scopeSlots[scopeIdx],
TypeHint: typeHint,
}
scope[name] = variable
p.scopeSlots[scopeIdx]++
return variable
}
func (p *Parser) declareLoopVariable(name string) {
p.scopes[len(p.scopes)-1][name] = true
func (p *Parser) declareLoopVariable(name string, typeHint TypeInfo) *Variable {
scope := p.scopes[len(p.scopes)-1]
scopeIdx := len(p.scopes) - 1
variable := &Variable{
Name: name,
ScopeDepth: p.currentDepth,
SlotIndex: p.scopeSlots[scopeIdx],
TypeHint: typeHint,
}
scope[name] = variable
p.scopeSlots[scopeIdx]++
return variable
}
func (p *Parser) lookupVariable(name string) *Variable {
for i := len(p.scopes) - 1; i >= 0; i-- {
if variable, exists := p.scopes[i][name]; exists {
return variable
}
}
return nil
}
// parseTypeHint parses optional type hint after colon, returns by value
@ -201,7 +259,7 @@ func (p *Parser) nextToken() {
// ParseProgram parses the entire program
func (p *Parser) ParseProgram() *Program {
program := &Program{}
program := &Program{Position: p.pos()}
program.Statements = []Statement{}
for !p.curTokenIs(EOF) {
@ -254,7 +312,7 @@ func (p *Parser) parseStatement() Statement {
// parseStructStatement parses struct definitions
func (p *Parser) parseStructStatement() *StructStatement {
stmt := &StructStatement{}
stmt := &StructStatement{Position: p.pos()}
if !p.expectPeek(IDENT) {
p.addError("expected struct name")
@ -289,7 +347,7 @@ func (p *Parser) parseStructStatement() *StructStatement {
return nil
}
field := StructField{Name: p.curToken.Literal}
field := StructField{Name: p.curToken.Literal, Position: p.pos()}
// Parse required type hint
field.TypeHint = p.parseTypeHint()
@ -327,6 +385,8 @@ func (p *Parser) parseStructStatement() *StructStatement {
// parseFunctionStatement handles both regular functions and methods
func (p *Parser) parseFunctionStatement() Statement {
pos := p.pos()
if !p.expectPeek(IDENT) {
p.addError("expected function name")
return nil
@ -358,7 +418,7 @@ func (p *Parser) parseFunctionStatement() Statement {
}
// Parse the function literal
funcLit := &FunctionLiteral{}
funcLit := &FunctionLiteral{Position: p.pos()}
funcLit.Parameters, funcLit.Variadic = p.parseFunctionParameters()
if !p.expectPeek(RPAREN) {
@ -373,7 +433,7 @@ func (p *Parser) parseFunctionStatement() Statement {
p.enterScope("function")
for _, param := range funcLit.Parameters {
p.declareVariable(param.Name)
p.declareVariable(param.Name, param.TypeHint)
}
funcLit.Body = p.parseBlockStatements(END)
p.exitScope()
@ -387,6 +447,7 @@ func (p *Parser) parseFunctionStatement() Statement {
StructID: structDef.ID,
MethodName: methodName,
Function: funcLit,
Position: pos,
}
}
@ -396,11 +457,13 @@ func (p *Parser) parseFunctionStatement() Statement {
return nil
}
return &ExpressionStatement{Expression: funcLit}
return &ExpressionStatement{Expression: funcLit, Position: pos}
}
// parseIdentifierStatement handles assignments and expression statements
func (p *Parser) parseIdentifierStatement() Statement {
pos := p.pos()
// Parse the left-hand side expression first
expr := p.ParseExpression(LOWEST)
if expr == nil {
@ -419,14 +482,23 @@ func (p *Parser) parseIdentifierStatement() Statement {
assignment := &Assignment{
Target: expr,
TypeHint: typeHint,
Position: pos,
}
// Validate assignment target and check if it's a declaration
switch target := expr.(type) {
case *Identifier:
assignment.IsDeclaration = !p.isVariableDeclared(target.Value)
if assignment.IsDeclaration {
p.declareVariable(target.Value)
if variable := p.lookupVariable(target.Value); variable != nil {
// Existing variable - resolve it
target.ScopeDepth = variable.ScopeDepth
target.SlotIndex = variable.SlotIndex
assignment.IsDeclaration = false
} else {
// New variable declaration - create and resolve
variable := p.declareVariable(target.Value, typeHint)
target.ScopeDepth = variable.ScopeDepth
target.SlotIndex = variable.SlotIndex
assignment.IsDeclaration = true
}
case *DotExpression, *IndexExpression:
assignment.IsDeclaration = false
@ -450,13 +522,13 @@ func (p *Parser) parseIdentifierStatement() Statement {
return assignment
} else {
// This is an expression statement
return &ExpressionStatement{Expression: expr}
return &ExpressionStatement{Expression: expr, Position: pos}
}
}
// parseEchoStatement parses echo statements
func (p *Parser) parseEchoStatement() *EchoStatement {
stmt := &EchoStatement{}
stmt := &EchoStatement{Position: p.pos()}
p.nextToken() // move past 'echo'
@ -475,11 +547,11 @@ func (p *Parser) parseBreakStatement() *BreakStatement {
p.addError("unexpected identifier")
return nil
}
return &BreakStatement{}
return &BreakStatement{Position: p.pos()}
}
func (p *Parser) parseExitStatement() *ExitStatement {
stmt := &ExitStatement{}
stmt := &ExitStatement{Position: p.pos()}
if p.canStartExpression(p.peekToken.Type) {
p.nextToken()
@ -494,7 +566,7 @@ func (p *Parser) parseExitStatement() *ExitStatement {
}
func (p *Parser) parseReturnStatement() *ReturnStatement {
stmt := &ReturnStatement{}
stmt := &ReturnStatement{Position: p.pos()}
if p.canStartExpression(p.peekToken.Type) {
p.nextToken()
@ -519,7 +591,7 @@ func (p *Parser) canStartExpression(tokenType TokenType) bool {
// Loop statement parsers
func (p *Parser) parseWhileStatement() *WhileStatement {
stmt := &WhileStatement{}
stmt := &WhileStatement{Position: p.pos()}
p.nextToken()
@ -547,6 +619,7 @@ func (p *Parser) parseWhileStatement() *WhileStatement {
}
func (p *Parser) parseForStatement() Statement {
pos := p.pos()
p.nextToken()
if !p.curTokenIs(IDENT) {
@ -554,20 +627,20 @@ func (p *Parser) parseForStatement() Statement {
return nil
}
firstVar := &Identifier{Value: p.curToken.Literal}
firstVar := &Identifier{Value: p.curToken.Literal, Position: p.pos()}
if p.peekTokenIs(ASSIGN) {
return p.parseNumericForStatement(firstVar)
return p.parseNumericForStatement(firstVar, pos)
} else if p.peekTokenIs(COMMA) || p.peekTokenIs(IN) {
return p.parseForInStatement(firstVar)
return p.parseForInStatement(firstVar, pos)
} else {
p.addError("expected '=', ',' or 'in' after for loop variable")
return nil
}
}
func (p *Parser) parseNumericForStatement(variable *Identifier) *ForStatement {
stmt := &ForStatement{Variable: variable}
func (p *Parser) parseNumericForStatement(variable *Identifier, pos Position) *ForStatement {
stmt := &ForStatement{Variable: variable, Position: pos}
if !p.expectPeek(ASSIGN) {
return nil
@ -613,7 +686,12 @@ func (p *Parser) parseNumericForStatement(variable *Identifier) *ForStatement {
p.nextToken()
p.enterScope("loop")
p.declareLoopVariable(variable.Value)
// Declare and resolve loop variable
loopVar := p.declareLoopVariable(variable.Value, NumberType)
variable.ScopeDepth = loopVar.ScopeDepth
variable.SlotIndex = loopVar.SlotIndex
variable.typeInfo = NumberType
stmt.Body = p.parseBlockStatements(END)
p.exitScope()
@ -625,8 +703,8 @@ func (p *Parser) parseNumericForStatement(variable *Identifier) *ForStatement {
return stmt
}
func (p *Parser) parseForInStatement(firstVar *Identifier) *ForInStatement {
stmt := &ForInStatement{}
func (p *Parser) parseForInStatement(firstVar *Identifier, pos Position) *ForInStatement {
stmt := &ForInStatement{Position: pos}
if p.peekTokenIs(COMMA) {
stmt.Key = firstVar
@ -638,7 +716,7 @@ func (p *Parser) parseForInStatement(firstVar *Identifier) *ForInStatement {
return nil
}
stmt.Value = &Identifier{Value: p.curToken.Literal}
stmt.Value = &Identifier{Value: p.curToken.Literal, Position: p.pos()}
} else {
stmt.Value = firstVar
}
@ -664,10 +742,20 @@ func (p *Parser) parseForInStatement(firstVar *Identifier) *ForInStatement {
p.nextToken()
p.enterScope("loop")
// Declare and resolve loop variables
if stmt.Key != nil {
p.declareLoopVariable(stmt.Key.Value)
keyVar := p.declareLoopVariable(stmt.Key.Value, AnyType)
stmt.Key.ScopeDepth = keyVar.ScopeDepth
stmt.Key.SlotIndex = keyVar.SlotIndex
stmt.Key.typeInfo = AnyType
}
p.declareLoopVariable(stmt.Value.Value)
valueVar := p.declareLoopVariable(stmt.Value.Value, AnyType)
stmt.Value.ScopeDepth = valueVar.ScopeDepth
stmt.Value.SlotIndex = valueVar.SlotIndex
stmt.Value.typeInfo = AnyType
stmt.Body = p.parseBlockStatements(END)
p.exitScope()
@ -681,7 +769,7 @@ func (p *Parser) parseForInStatement(firstVar *Identifier) *ForInStatement {
// parseIfStatement parses if statements
func (p *Parser) parseIfStatement() *IfStatement {
stmt := &IfStatement{}
stmt := &IfStatement{Position: p.pos()}
p.nextToken()
@ -705,7 +793,7 @@ func (p *Parser) parseIfStatement() *IfStatement {
stmt.Body = p.parseBlockStatements(ELSEIF, ELSE, END)
for p.curTokenIs(ELSEIF) {
elseif := ElseIfClause{}
elseif := ElseIfClause{Position: p.pos()}
p.nextToken()
@ -793,11 +881,25 @@ func (p *Parser) ParseExpression(precedence Precedence) Expression {
// Expression parsing functions
func (p *Parser) parseIdentifier() Expression {
return &Identifier{Value: p.curToken.Literal}
ident := &Identifier{
Value: p.curToken.Literal,
ScopeDepth: -1,
SlotIndex: -1,
Position: p.pos(),
}
// Resolve variable if it exists
if variable := p.lookupVariable(ident.Value); variable != nil {
ident.ScopeDepth = variable.ScopeDepth
ident.SlotIndex = variable.SlotIndex
ident.typeInfo = variable.TypeHint
}
return ident
}
func (p *Parser) parseNumberLiteral() Expression {
lit := &NumberLiteral{}
lit := &NumberLiteral{Position: p.pos()}
literal := p.curToken.Literal
var value float64
@ -853,20 +955,21 @@ func (p *Parser) parseNumberLiteral() Expression {
}
func (p *Parser) parseStringLiteral() Expression {
return &StringLiteral{Value: p.curToken.Literal}
return &StringLiteral{Value: p.curToken.Literal, Position: p.pos()}
}
func (p *Parser) parseBooleanLiteral() Expression {
return &BooleanLiteral{Value: p.curTokenIs(TRUE)}
return &BooleanLiteral{Value: p.curTokenIs(TRUE), Position: p.pos()}
}
func (p *Parser) parseNilLiteral() Expression {
return &NilLiteral{}
return &NilLiteral{Position: p.pos()}
}
func (p *Parser) parsePrefixExpression() Expression {
expression := &PrefixExpression{
Operator: p.curToken.Literal,
Position: p.pos(),
}
p.nextToken()
@ -904,6 +1007,7 @@ func (p *Parser) parseGroupedExpression() Expression {
// parseParenthesizedAssignment parses assignment expressions in parentheses
func (p *Parser) parseParenthesizedAssignment() Expression {
pos := p.pos()
target := p.parseIdentifier()
if !p.expectPeek(ASSIGN) {
@ -927,13 +1031,22 @@ func (p *Parser) parseParenthesizedAssignment() Expression {
Target: target,
Value: value,
IsExpression: true,
Position: pos,
}
// Handle variable declaration for assignment expressions
if ident, ok := target.(*Identifier); ok {
assignExpr.IsDeclaration = !p.isVariableDeclared(ident.Value)
if assignExpr.IsDeclaration {
p.declareVariable(ident.Value)
if variable := p.lookupVariable(ident.Value); variable != nil {
// Existing variable
ident.ScopeDepth = variable.ScopeDepth
ident.SlotIndex = variable.SlotIndex
assignExpr.IsDeclaration = false
} else {
// New variable declaration
newVar := p.declareVariable(ident.Value, UnknownType)
ident.ScopeDepth = newVar.ScopeDepth
ident.SlotIndex = newVar.SlotIndex
assignExpr.IsDeclaration = true
}
}
@ -941,7 +1054,7 @@ func (p *Parser) parseParenthesizedAssignment() Expression {
}
func (p *Parser) parseFunctionLiteral() Expression {
fn := &FunctionLiteral{}
fn := &FunctionLiteral{Position: p.pos()}
if !p.expectPeek(LPAREN) {
p.addError("expected '(' after 'fn'")
@ -961,10 +1074,21 @@ func (p *Parser) parseFunctionLiteral() Expression {
p.nextToken()
p.enterScope("function")
// Declare function parameters as resolved variables
for _, param := range fn.Parameters {
p.declareVariable(param.Name)
p.declareVariable(param.Name, param.TypeHint)
param.Position = Position{Line: p.curToken.Line, Column: p.curToken.Column}
}
fn.Body = p.parseBlockStatements(END)
// Calculate function metadata before exiting scope
scopeIdx := len(p.scopes) - 1
fn.LocalCount = p.scopeSlots[scopeIdx]
fn.UpvalueCount = 0 // TODO: Calculate captured variables
fn.MaxStackDepth = 0 // TODO: Calculate max expression depth
p.exitScope()
if !p.curTokenIs(END) {
@ -996,7 +1120,7 @@ func (p *Parser) parseFunctionParameters() ([]FunctionParameter, bool) {
return nil, false
}
param := FunctionParameter{Name: p.curToken.Literal}
param := FunctionParameter{Name: p.curToken.Literal, Position: p.pos()}
// Check for type hint
param.TypeHint = p.parseTypeHint()
@ -1020,7 +1144,7 @@ func (p *Parser) parseFunctionParameters() ([]FunctionParameter, bool) {
}
func (p *Parser) parseTableLiteral() Expression {
table := &TableLiteral{}
table := &TableLiteral{Position: p.pos()}
table.Pairs = make([]TablePair, 0, 4) // Pre-allocate
if p.peekTokenIs(RBRACE) {
@ -1036,13 +1160,13 @@ func (p *Parser) parseTableLiteral() Expression {
return nil
}
pair := TablePair{}
pair := TablePair{Position: p.pos()}
if (p.curTokenIs(IDENT) || p.curTokenIs(STRING)) && p.peekTokenIs(ASSIGN) {
if p.curTokenIs(IDENT) {
pair.Key = &Identifier{Value: p.curToken.Literal}
pair.Key = &Identifier{Value: p.curToken.Literal, Position: p.pos()}
} else {
pair.Key = &StringLiteral{Value: p.curToken.Literal}
pair.Key = &StringLiteral{Value: p.curToken.Literal, Position: p.pos()}
}
p.nextToken()
p.nextToken()
@ -1089,6 +1213,8 @@ func (p *Parser) parseTableLiteral() Expression {
// parseStructConstructor handles struct constructor calls
func (p *Parser) parseStructConstructor(left Expression) Expression {
pos := p.pos()
ident, ok := left.(*Identifier)
if !ok {
return p.parseTableLiteralFromBrace()
@ -1105,6 +1231,7 @@ func (p *Parser) parseStructConstructor(left Expression) Expression {
StructID: structDef.ID,
Fields: make([]TablePair, 0, 4),
typeInfo: TypeInfo{Type: TypeStruct, StructID: structDef.ID, Inferred: true},
Position: pos,
}
if p.peekTokenIs(RBRACE) {
@ -1120,13 +1247,13 @@ func (p *Parser) parseStructConstructor(left Expression) Expression {
return nil
}
pair := TablePair{}
pair := TablePair{Position: p.pos()}
if (p.curTokenIs(IDENT) || p.curTokenIs(STRING)) && p.peekTokenIs(ASSIGN) {
if p.curTokenIs(IDENT) {
pair.Key = &Identifier{Value: p.curToken.Literal}
pair.Key = &Identifier{Value: p.curToken.Literal, Position: p.pos()}
} else {
pair.Key = &StringLiteral{Value: p.curToken.Literal}
pair.Key = &StringLiteral{Value: p.curToken.Literal, Position: p.pos()}
}
p.nextToken()
p.nextToken()
@ -1172,7 +1299,7 @@ func (p *Parser) parseStructConstructor(left Expression) Expression {
}
func (p *Parser) parseTableLiteralFromBrace() Expression {
table := &TableLiteral{}
table := &TableLiteral{Position: p.pos()}
table.Pairs = make([]TablePair, 0, 4)
if p.peekTokenIs(RBRACE) {
@ -1188,13 +1315,13 @@ func (p *Parser) parseTableLiteralFromBrace() Expression {
return nil
}
pair := TablePair{}
pair := TablePair{Position: p.pos()}
if (p.curTokenIs(IDENT) || p.curTokenIs(STRING)) && p.peekTokenIs(ASSIGN) {
if p.curTokenIs(IDENT) {
pair.Key = &Identifier{Value: p.curToken.Literal}
pair.Key = &Identifier{Value: p.curToken.Literal, Position: p.pos()}
} else {
pair.Key = &StringLiteral{Value: p.curToken.Literal}
pair.Key = &StringLiteral{Value: p.curToken.Literal, Position: p.pos()}
}
p.nextToken()
p.nextToken()
@ -1243,6 +1370,7 @@ func (p *Parser) parseInfixExpression(left Expression) Expression {
expression := &InfixExpression{
Left: left,
Operator: p.curToken.Literal,
Position: p.pos(),
}
precedence := p.curPrecedence()
@ -1258,6 +1386,8 @@ func (p *Parser) parseInfixExpression(left Expression) Expression {
}
func (p *Parser) parseDotExpression(left Expression) Expression {
pos := p.pos()
if !p.expectPeekIdent() {
p.addError("expected identifier after '.'")
return nil
@ -1266,11 +1396,12 @@ func (p *Parser) parseDotExpression(left Expression) Expression {
return &DotExpression{
Left: left,
Key: p.curToken.Literal,
Position: pos,
}
}
func (p *Parser) parseCallExpression(fn Expression) Expression {
call := &CallExpression{Function: fn}
call := &CallExpression{Function: fn, Position: p.pos()}
call.Arguments = p.parseExpressionList(RPAREN)
return call
}
@ -1300,6 +1431,7 @@ func (p *Parser) parseExpressionList(end TokenType) []Expression {
}
func (p *Parser) parseIndexExpression(left Expression) Expression {
pos := p.pos()
p.nextToken()
index := p.ParseExpression(LOWEST)
@ -1316,6 +1448,7 @@ func (p *Parser) parseIndexExpression(left Expression) Expression {
return &IndexExpression{
Left: left,
Index: index,
Position: pos,
}
}
@ -1381,7 +1514,7 @@ func (p *Parser) noPrefixParseFnError(t TokenType) {
switch t {
case ASSIGN:
message = "unexpected assignment operator, missing left-hand side identifier"
case PLUS, MINUS, STAR, SLASH:
case PLUS, MINUS, STAR, SLASH, MOD:
message = fmt.Sprintf("unexpected operator '%s', missing left operand", tokenTypeString(t))
case RPAREN:
message = "unexpected closing parenthesis"
@ -1446,6 +1579,8 @@ func tokenTypeString(t TokenType) string {
return "*"
case SLASH:
return "/"
case MOD:
return "%"
case DOT:
return "."
case COLON:

43
parser/tests/parser_test.go
View File

@ -325,3 +325,46 @@ echo {result = x}`
t.Fatalf("expected TableLiteral in echo, got %T", echo2.Value)
}
}
func TestModuloOperator(t *testing.T) {
input := `result = 10 % 3`
l := parser.NewLexer(input)
p := parser.NewParser(l)
program := p.ParseProgram()
checkParserErrors(t, p)
if len(program.Statements) != 1 {
t.Fatalf("expected 1 statement, got %d", len(program.Statements))
}
stmt, ok := program.Statements[0].(*parser.Assignment)
if !ok {
t.Fatalf("expected AssignStatement, got %T", program.Statements[0])
}
infix, ok := stmt.Value.(*parser.InfixExpression)
if !ok {
t.Fatalf("expected InfixExpression, got %T", stmt.Value)
}
if infix.Operator != "%" {
t.Errorf("expected operator '%%', got %s", infix.Operator)
}
leftLit, ok := infix.Left.(*parser.NumberLiteral)
if !ok {
t.Fatalf("expected NumberLiteral on left, got %T", infix.Left)
}
if leftLit.Value != 10.0 {
t.Errorf("expected left value 10, got %f", leftLit.Value)
}
rightLit, ok := infix.Right.(*parser.NumberLiteral)
if !ok {
t.Fatalf("expected NumberLiteral on right, got %T", infix.Right)
}
if rightLit.Value != 3.0 {
t.Errorf("expected right value 3, got %f", rightLit.Value)
}
}

4
parser/token.go
View File

@ -18,6 +18,7 @@ const (
MINUS // -
STAR // *
SLASH // /
MOD // %
DOT // .
// Comparison operators
@ -85,7 +86,7 @@ const (
EQUALS // ==, !=
LESSGREATER // >, <, >=, <=
SUM // +, -
PRODUCT // *, /
PRODUCT // *, /, %
PREFIX // -x, not x
MEMBER // table[key], table.key
CALL // function()
@ -105,6 +106,7 @@ var precedences = map[TokenType]Precedence{
MINUS: SUM,
SLASH: PRODUCT,
STAR: PRODUCT,
MOD: PRODUCT,
DOT: MEMBER,
LBRACKET: MEMBER,
LPAREN: CALL,

5
parser/types.go
View File

@ -521,7 +521,7 @@ func (ti *TypeInferrer) inferInfixExpression(infix *InfixExpression) TypeInfo {
var resultType TypeInfo
switch infix.Operator {
case "+", "-", "*", "/":
case "+", "-", "*", "/", "%":
if !ti.isNumericType(leftType) || !ti.isNumericType(rightType) {
ti.addError(fmt.Sprintf("arithmetic operator '%s' requires numeric operands", infix.Operator), infix)
}
@ -636,8 +636,11 @@ func (ti *TypeInferrer) exitScope() {
}
func (ti *TypeInferrer) addError(message string, node Node) {
pos := node.Pos()
ti.errors = append(ti.errors, TypeError{
Message: message,
Line: pos.Line,
Column: pos.Column,
Node: node,
})
}