1 files changed, 28 insertions(+), 9 deletions(-)

M internal/bootstrap/scanner/scanner.go

@@ 5,6 5,7 @@ package scanner
 
 import (
 	"fmt"
+	"go/scanner"
 	"unicode"
 	"unicode/utf8"
 


@@ 18,13 19,17 @@ const (
 	eof = -1
 )
 
+// ErrorList is a list of errors.
+type ErrorList struct{ scanner.ErrorList }
+
 // A Scanner holds the scanner's internal state while processing a
 // given input. It can be allocated as part of another data structure
 // but must be initialized via Init before use.
 type Scanner struct {
-	// immutable fields
+	// immutable fields outside of Init calls
 	file *token.File
 	src  []byte
+	errs *ErrorList
 
 	// mutable state
 	rn      rune // current rune


@@ 40,7 45,7 @@ type Scanner struct {
 // size.
 func (s *Scanner) Init(file *token.File, src []byte) {
 	if file.Size() != len(src) {
-		panic(fmt.Sprintf("file size does not match input size: %d bytes vs %d bytes", file.Size(), len(src)))
+		panic(fmt.Sprintf("%s: file size does not match input size: %d bytes vs %d bytes", file.Name(), file.Size(), len(src)))
 	}
 
 	s.file = file


@@ 78,14 83,14 @@ func (s *Scanner) advance() {
 	rn, sz := rune(s.src[s.nextpos]), 1
 	switch {
 	case rn == 0:
-		// TODO: error, illegal NUL character
+		s.error(s.curpos, "illegal character NUL")
 	case rn >= utf8.RuneSelf:
 		// not ascii
 		rn, sz = utf8.DecodeRune(s.src[s.nextpos:])
 		if rn == utf8.RuneError && sz == 1 {
-			// TODO: illegal utf8 encoding
+			s.error(s.curpos, "illegal UTF-8 encoding")
 		} else if rn == bom && s.curpos > 0 {
-			// TODO: illegal BOM mark
+			s.error(s.curpos, "illegal byte-order mark")
 		}
 	}
 


@@ 93,6 98,11 @@ func (s *Scanner) advance() {
 	s.rn = rn
 }
 
+func (s *Scanner) error(offset int, msg string) {
+	fpos := s.file.Position(s.file.Pos(offset))
+	s.errs.Add(fpos, msg)
+}
+
 func (s *Scanner) peek() byte {
 	if s.nextpos < len(s.src) {
 		return s.src[s.nextpos]


@@ 144,7 154,7 @@ func (s *Scanner) Scan() (pos token.Pos, tok token.Token, lit string) {
 
 		case '<':
 			if s.rn != '-' {
-				// TODO: Illegal, report error
+				s.error(s.curpos-1, fmt.Sprintf("incomplete arrow symbol: illegal character %#U", s.rn))
 				break
 			}
 			s.advance()


@@ 169,10 179,18 @@ func (s *Scanner) Scan() (pos token.Pos, tok token.Token, lit string) {
 			lit = s.scanCoderef()
 
 		case '$':
+			if s.rn != '{' {
+				s.error(s.curpos-1, fmt.Sprintf("invalid state coderef: illegal character %#U", s.rn))
+				break
+			}
 			tok = token.StateCoderef
 			lit = s.scanCoderef()
 
 		case '@':
+			if s.rn != '{' {
+				s.error(s.curpos-1, fmt.Sprintf("invalid state coderef: illegal character %#U", s.rn))
+				break
+			}
 			tok = token.ActionCoderef
 			lit = s.scanCoderef()
 


@@ 182,7 200,7 @@ func (s *Scanner) Scan() (pos token.Pos, tok token.Token, lit string) {
 		default:
 			// s.advance already reports bom, ignore it here
 			if rn != bom {
-				// TODO: report illegal char
+				s.error(s.file.Offset(pos.Pos), fmt.Sprintf("illegal character %#U", rn))
 			}
 			tok = token.Illegal
 			lit = string(rn)


@@ 212,6 230,7 @@ func (s *Scanner) scanCharClass() string {
 
 func (s *Scanner) scanCoderef() string {
 	// when called, s.rn is on the start '{'
+	start := s.curpos
 	s.advance()
 
 	s.skipWhitespace()


@@ 221,10 240,10 @@ func (s *Scanner) scanCoderef() string {
 	if s.rn == '}' {
 		s.advance()
 	} else {
-		// TODO: error, unclosed coderef
+		s.error(start, "coderef not terminated")
 	}
 	if ident == "" {
-		// TODO: error, empty coderef
+		s.error(start, "coderef missing identifier")
 	}
 	return ident
 }