func (s *Scanner) scanNumber() (token.Pos, string) {
start := s.off
for unicode.IsDigit(s.ch) {
s.next()
}
return token.Pos(start), s.str[start:s.off]
}
func (s *Scanner) scanComment() (token.Pos, string) {
start := s.off - 1
for s.ch != '\n' && s.off < len(s.str) {
s.next()
}
return token.Pos(start), s.str[start:s.off]
}
func (s *Scanner) scanIdentifier() (token.Pos, string) {
start := s.off
for unicode.IsDigit(s.ch) || isAlpha(s.ch) || s.ch == '_' || s.ch == '-' {
s.next()
}
return token.Pos(start), s.str[start:s.off]
}
func ParseFile(f *token.File, str string) *ast.File {
if f.Size() != len(str) {
fmt.Println("File size does not match string length.")
return nil
}
root := ast.NewFile(token.Pos(1), token.Pos(len(str)+1))
p := new(parser)
p.init(f, str)
p.topScope = root.Scope
p.curScope = root.Scope
n := p.parse()
for p.file.NumErrors() < 10 && p.tok != token.EOF {
root.Nodes = append(root.Nodes, n)
p.next()
n = p.parse()
}
if p.topScope != p.curScope {
panic("Imbalanced scope!")
}
return root
}
func (s *Scanner) next() {
s.off = s.roff
if s.off < len(s.str) {
r, n := utf8.DecodeRuneInString(s.str[s.off:])
if r == utf8.RuneError {
s.file.AddError(s.file.Base()+token.Pos(s.off), "Invalid UTF8 string!")
return
}
s.ch = r
s.roff += n
} else {
s.ch = 0
}
if s.ch == '\n' {
s.file.AddLine(s.off)
}
//s.roff++
}
func (s *Scanner) Scan() (tok token.Token, pos token.Pos, lit string) {
s.skipWhitespace()
if isAlpha(s.ch) {
pos, lit = s.scanIdentifier()
tok = token.Lookup(lit)
return
}
if unicode.IsDigit(s.ch) {
pos, lit = s.scanNumber()
return token.NUMBER, pos, lit
}
ch := s.ch
pos = token.Pos(s.off)
s.next()
switch ch {
case '+':
tok = token.ADD
case '-':
if unicode.IsDigit(s.ch) { // is '-' a unary operator for a number?
pos, lit = s.scanNumber()
pos, lit, tok = pos-1, string('-')+lit, token.NUMBER
return
}
tok = token.SUB
case '*':
tok = token.MUL
case '/':
tok = token.DIV
case '%':
tok = token.MOD
case '(':
tok = token.LPAREN
case ')':
tok = token.RPAREN
case ';':
pos, lit = s.scanComment()
tok = token.COMMENT
return
case '=':
tok = token.EQ
case '<':
switch s.ch {
case '=':
lit = string(ch) + string(s.ch)
tok = token.LTE
s.next()
return
case '>':
lit = string(ch) + string(s.ch)
tok = token.NEQ
s.next()
return
}
tok = token.LT
case '>':
if s.ch == '=' {
lit = string(ch) + string(s.ch)
tok = token.GTE
s.next()
return
}
tok = token.GT
case '"':
lit = s.scanString()
tok = token.STRING
return
default:
if s.off >= len(s.str) {
tok = token.EOF
pos = token.Pos(s.off)
return
}
}
lit = string(ch)
return
}
func (n *Number) End() token.Pos { return n.Num + token.Pos(len(n.Lit)) }
func (i *Identifier) End() token.Pos { return i.Id + token.Pos(len(i.Lit)) }
func (s *String) End() token.Pos { return s.Str + token.Pos(len(s.Lit)) }