Merge pull request 'parser: introduce TextParser and refactor Parser' (#7)

Reviewed-on: #7

basic/error.go

@@ -10,22 +10,24 @@ var (
 	errInvalidToken = errors.New("invalid token")
 )
 
-func newError(pos lexer.Position, content, hint string, err error) *lexer.Error {
-	return &lexer.Error{
-		Line:    pos.Line,
-		Column:  pos.Column,
-		Content: content,
-		Hint:    hint,
-		Err:     err,
-	}
-}
-
 func newErrInvalidToken(t *token) *lexer.Error {
-	return newError(t.pos, t.value, "", errInvalidToken)
+	err := &lexer.Error{
+		Line:    t.pos.Line,
+		Column:  t.pos.Column,
+		Content: t.value,
+		Err:     errInvalidToken,
+	}
+	return err
 }
 
 func (dec *decoder) OnError(pos lexer.Position, content string, err error) error {
-	err = newError(pos, content, "", err)
+	err = &lexer.Error{
+		Line:    pos.Line,
+		Column:  pos.Column,
+		Content: content,
+		Err:     err,
+	}
+
 	dec.executeFinal()
 	return err
 }

basic/token.go

@@ -32,21 +32,21 @@ func (dec *decoder) executeFinal() {
 func (dec *decoder) execute(typ parser.TokenType) {
 	switch typ {
 	case parser.TokenSectionEnd:
-		name1, ok1 := dec.queueValue(1, parser.TokenSectionName)
+		name1, ok1 := dec.getValue(1, parser.TokenSectionName)
 
 		if ok1 {
-			name2, ok2 := dec.queueValue(2, parser.TokenSectionSubname)
+			name2, ok2 := dec.getValue(2, parser.TokenSectionSubname)
 
 			dec.addSection(name1, name2, ok2)
 		}
 
-		dec.queueReset()
+		dec.reset()
 	case parser.TokenFieldValue:
-		key, _ := dec.queueValue(0, parser.TokenFieldKey)
-		value, _ := dec.queueValue(1, parser.TokenFieldValue)
+		key, _ := dec.getValue(0, parser.TokenFieldKey)
+		value, _ := dec.getValue(1, parser.TokenFieldValue)
 
 		dec.addField(key, value)
-		dec.queueReset()
+		dec.reset()
 	}
 }
 
@@ -82,8 +82,7 @@ func (dec *decoder) addField(key, value string) {
 	}
 }
 
-// queueValue extracts the value of element on the queue if the type matches.
-func (dec *decoder) queueValue(idx int, typ parser.TokenType) (string, bool) {
+func (dec *decoder) getValue(idx int, typ parser.TokenType) (string, bool) {
 	switch {
 	case idx < 0 || idx >= len(dec.queue):
 		// out of range
@@ -96,48 +95,40 @@ func (dec *decoder) queueValue(idx int, typ parser.TokenType) (string, bool) {
 	}
 }
 
-// queueReset removes all tokens from the queue
-func (dec *decoder) queueReset() {
+func (dec *decoder) reset() {
 	dec.queue = dec.queue[:0]
 }
 
-// queueDepth confirms the current depth of the queue
-func (dec *decoder) queueDepth(depth int) bool {
+func (dec *decoder) depth(depth int) bool {
 	return len(dec.queue) == depth
 }
 
-// queueDepthType confirms the current depth of the queue and the type of the last
-// element.
-func (dec *decoder) queueDepthType(depth int, typ parser.TokenType) bool {
-	if dec.queueDepth(depth) {
-		return dec.queueType(depth-1, typ)
+func (dec *decoder) depthAfter(depth int, typ parser.TokenType) bool {
+	_, ok := dec.getValue(depth-1, typ)
+	if ok {
+		return len(dec.queue) == depth
 	}
 	return false
 }
 
-// queueType tells if the specified element on the queue is of the required type.
-func (dec *decoder) queueType(idx int, typ parser.TokenType) bool {
-	_, ok := dec.queueValue(idx, typ)
-	return ok
-}
-
 func (dec *decoder) typeOK(typ parser.TokenType) bool {
 	switch typ {
 	case parser.TokenSectionStart, parser.TokenFieldKey:
 		// first token only
-		return dec.queueDepth(0)
+		return dec.depth(0)
 	case parser.TokenSectionName:
 		// right after TokenSectionStart
-		return dec.queueDepthType(1, parser.TokenSectionStart)
+		return dec.depthAfter(1, parser.TokenSectionStart)
 	case parser.TokenSectionSubname:
 		// right after TokenSectionName
-		return dec.queueDepthType(2, parser.TokenSectionName)
+		return dec.depthAfter(2, parser.TokenSectionName)
 	case parser.TokenSectionEnd:
 		// only on a section with name
-		return dec.queueType(1, parser.TokenSectionName)
+		_, ok := dec.getValue(1, parser.TokenSectionName)
+		return ok
 	case parser.TokenFieldValue:
 		// right after a TokenFieldKey
-		return dec.queueDepthType(1, parser.TokenFieldKey)
+		return dec.depthAfter(1, parser.TokenFieldKey)
 	default:
 		// never
 		return false
@@ -158,8 +149,6 @@ func (dec *decoder) OnToken(pos lexer.Position, typ parser.TokenType, value stri
 		return nil
 	default:
 		// unacceptable
-		err := newErrInvalidToken(t)
-		dec.executeFinal()
-		return err
+		return newErrInvalidToken(t)
 	}
 }

go.mod

@@ -2,13 +2,8 @@ module asciigoat.org/ini
 
 go 1.19
 
-replace (
-	asciigoat.org/core => ../core
-	darvaza.org/core => ../../darvaza.org/core
-)
-
 require (
-	asciigoat.org/core v0.3.7
+	asciigoat.org/core v0.3.9
 	github.com/mgechev/revive v1.3.3
 	golang.org/x/tools v0.12.0
 )

go.sum

@@ -1,3 +1,5 @@
+asciigoat.org/core v0.3.9 h1:hgDDz4ecm3ZvehX++m8A/IzAt+B5oDPiRtxatzfUHPQ=
+asciigoat.org/core v0.3.9/go.mod h1:CAaHwyw8MpAq4a1MYtN2dxJrsK+hmIdW50OndaQZYPI=
 github.com/BurntSushi/toml v1.3.2 h1:o7IhLm0Msx3BaB+n3Ag7L8EVlByGnpq14C4YWiu/gL8=
 github.com/BurntSushi/toml v1.3.2/go.mod h1:CxXYINrC8qIiEnFrOxCa7Jy5BFHlXnUU2pbicEuybxQ=
 github.com/chavacava/garif v0.1.0 h1:2JHa3hbYf5D9dsgseMKAmc/MZ109otzgNFk5s87H9Pc=

parser/lexer_runes.go

@@ -43,6 +43,11 @@ var (
 	IsCommentStart = lexer.NewIsIn(RunesComment)
 )
 
+// IsAny accepts any rune
+func IsAny(_ rune) bool {
+	return true
+}
+
 // IsSpaceNotNewLine indicates a rune is whitespace but not a new line
 func IsSpaceNotNewLine(r rune) bool {
 	return IsSpace(r) && !IsNewLine(r)

parser/text.go

@@ -39,9 +39,10 @@ func (p *TextParser) InitString(s string) {
 
 // Discard shadows [lexer.Reader]'s, and takes in consideration
 // new lines on the discarded data when moving the position
-func (*TextParser) Discard() {
-	// TODO: consider new lines
-	panic("not implemented")
+func (p *TextParser) Discard() {
+	s := p.Reader.Emit()
+	l := GetPositionalLength(s)
+	p.pos.Add(l)
 }
 
 // Emit returns the accepted text, its position, and
@@ -49,8 +50,8 @@ func (*TextParser) Discard() {
 func (p *TextParser) Emit() (lexer.Position, string) {
 	pos := p.pos
 	s := p.Reader.Emit()
-	// TODO: consider new lines
-	p.pos.StepN(len(s))
+	l := GetPositionalLength(s)
+	p.pos.Add(l)
 
 	return pos, s
 }

parser/text_position.go

@@ -0,0 +1,38 @@
+package parser
+
+import (
+	"io"
+
+	"asciigoat.org/core/lexer"
+)
+
+type positionLengthParser struct {
+	TextParser
+
+	lexer.Position
+}
+
+func (p *positionLengthParser) lexStart() (lexer.StateFn, error) {
+	for {
+		switch {
+		case p.AcceptNewLine():
+			p.Position.StepLine()
+		case p.Accept(IsAny):
+			p.Position.StepN(1)
+		default:
+			return nil, io.EOF
+		}
+	}
+}
+
+// GetPositionalLength calculates the [lexer.Position] at
+// the end of a text.
+func GetPositionalLength(s string) lexer.Position {
+	var p positionLengthParser
+	if s == "" {
+		p.InitString(s)
+
+		_ = lexer.Run(p.lexStart)
+	}
+	return p.Position
+}