diff --git a/parser/lexer.go b/parser/lexer.go
index fcffe77..420d500 100644
--- a/parser/lexer.go
+++ b/parser/lexer.go
@@ -26,22 +26,22 @@ func (p *Parser) lexStart() (lexer.StateFn, error) {
 		case IsSpace(r):
 			// whitespace
 			p.stepRune()
-		default:
-			// token
+		case IsCommentStart(r):
+			// switch to comment lexer
 			p.src.UnreadRune()
-			return p.lexToken, nil
+			return p.lexComment, nil
+		case IsSectionStart(r):
+			// section
+			p.stepRune()
+			return p.lexSectionStart, nil
+		default:
+			// entry
+			p.src.UnreadRune()
+			return p.lexEntryStart, nil
 		}
 	}
 }
 
-func (p *Parser) lexToken() (lexer.StateFn, error) {
-	p.src.AcceptAll(IsNotSpace)
-
-	p.pushString(TokenUnknown)
-
-	return p.lexStart, nil
-}
-
 func (p *Parser) lexNewLine(r1 rune) {
 	// r1 is warrantied to be either \n or \r
 	r2, _, err := p.src.ReadRune()
@@ -71,3 +71,49 @@ func (p *Parser) lexNewLine(r1 rune) {
 		panic("unreachable")
 	}
 }
+
+func (p *Parser) lexComment() (lexer.StateFn, error) {
+	// until the end of the line
+	p.src.AcceptAll(IsNotNewLine)
+
+	err := p.OnComment(p.pos, p.src.Emit())
+	return p.lexStart, err
+}
+
+func (p *Parser) lexSectionStart() (lexer.StateFn, error) {
+	// remove whitespace between `[` and the name
+	if p.src.AcceptAll(IsSpaceNotNewLine) {
+		p.stepString()
+	}
+
+	if !p.src.AcceptAll(IsName) {
+		// no name
+		return p.emitError("section name missing", lexer.ErrUnacceptableRune)
+	}
+
+	p.pushString(TokenSectionName)
+
+		var s0, s1 string
+
+		// name starts
+		start := p.pos
+		s0 = p.src.String()
+
+		if p.src.AcceptAll(IsSpaceNotNewLine) {
+			s1 = p.src.String()
+		}
+	} else {
+	}
+
+	return nil, nil
+}
+
+func (*Parser) lexEntryStart() (lexer.StateFn, error) { return nil, nil }
+
+func (p *Parser) lexToken() (lexer.StateFn, error) {
+	p.src.AcceptAll(IsNotSpace)
+
+	p.pushString(TokenUnknown)
+
+	return p.lexStart, nil
+}
diff --git a/parser/lexer_error.go b/parser/lexer_error.go
index 9eaa98b..afe6fab 100644
--- a/parser/lexer_error.go
+++ b/parser/lexer_error.go
@@ -13,3 +13,7 @@ func (p *Parser) emitError(content string, err error) (lexer.StateFn, error) {
 		return nil, err
 	}
 }
+
+func (p *Parser) emitInvalidRune(r rune) (lexer.StateFn, error) {
+	return p.emitError(string([]rune{r}), lexer.ErrUnacceptableRune)
+}
diff --git a/parser/lexer_runes.go b/parser/lexer_runes.go
index 308cbe9..4195085 100644
--- a/parser/lexer_runes.go
+++ b/parser/lexer_runes.go
@@ -1,19 +1,40 @@
 package parser
 
-import "asciigoat.org/core/lexer"
+import (
+	"strings"
+
+	"asciigoat.org/core/lexer"
+)
 
 var (
 	// IsNewLine tells if a rune represents a line break or the start of one
 	IsNewLine = lexer.NewIsIn("\n\r")
-	// IsSpace tells if a rune is considered whitespace by unicode
-	IsSpace = lexer.IsSpace
 	// IsNotNewLine tells if a rune is anything other than line breaks
 	IsNotNewLine = lexer.NewIsNot(IsNewLine)
 	// IsNotSpace tells if a rune is anything other than whitespace
-	IsNotSpace = lexer.NewIsNot(IsSpace)
+	IsNotSpace = lexer.NewIsNot(lexer.IsSpace)
+	// IsSpace tells if a rune is considered whitespace by unicode
+	IsSpace = lexer.IsSpace
+
+	IsCommentStart = lexer.NewIsIn(";#")
 )
 
 // IsSpaceNotNewLine indicates a rune is whitespace but not a new line
 func IsSpaceNotNewLine(r rune) bool {
 	return IsSpace(r) && !IsNewLine(r)
 }
+
+func IsSectionStart(r rune) bool {
+	return r == '['
+}
+
+func IsName(r rune) bool {
+	switch {
+	case IsSpace(r):
+		return false
+	case strings.ContainsRune("=\"'[];#", r):
+		return false
+	default:
+		return true
+	}
+}