parser: SplitCommaArray [WIP]

Signed-off-by: Alejandro Mery <amery@jpi.io>

parser/comma_array.go

@@ -1,8 +1,6 @@
 package parser
 
 import (
-	"strings"
-
 	"asciigoat.org/core/lexer"
 )
 
@@ -48,13 +46,18 @@ func (p *commaArrayParser) lexWord() (lexer.StateFn, error) {
 			_, s := p.Emit()
 			p.out = append(p.out, s)
 			return nil, err
-		case r == ',':
-			// done
-			_, s := p.Emit()
-			// remove comma, trim and append to output
-			s = strings.TrimRightFunc(s[:len(s)-1], IsSpace)
-			p.out = append(p.out, s)
-			return p.lexStart, nil
+		case r == RuneEscape:
+			// escaped
+			r2, _, err := p.ReadRune()
+			switch {
+			case err != nil:
+				// incomplete
+				return nil, NewErrIncompleteEscaped(&p.TextParser)
+			case IsNewLine(r2):
+				// escaped new line
+				p.UnreadRune()
+				p.AcceptNewLine()
+			}
 		}
 	}
 }
@@ -76,8 +79,7 @@ func (p *commaArrayParser) Run() ([]string, error) {
 	return p.out, err
 }
 
-// SplitCommaArray splits comma separated strings, removing whitespace
-// and respecting quoted literals.
+// SplitCommaArray splits
 func SplitCommaArray(s string) ([]string, error) {
 	if s != "" {
 		var p commaArrayParser

parser/error.go

@@ -35,6 +35,13 @@ func NewErrIncompleteQuotedString(p *TextParser) *lexer.Error {
 	return newErrIncomplete(p, "incomplete quoted string")
 }
 
+// NewErrIncompleteEscaped returns a [lexer.Error]
+// indicating the text being parsed wasn't correctly
+// terminated
+func NewErrIncompleteEscaped(p *TextParser) *lexer.Error {
+	return newErrIncomplete(p, "incomplete escaped string")
+}
+
 func newErrIncomplete(p *TextParser, hint string) *lexer.Error {
 	pos, s := p.Emit()
 	pos.Add(GetPositionalLength(s))

parser/text_quoted.go

@@ -1,13 +1,11 @@
 package parser
 
 import (
-	"strings"
-
 	"asciigoat.org/core/lexer"
 )
 
 // AcceptQuotedString consumes a quoted string from the source
-// and returns it unquoted and unescaped
+// and returns it unquoted.
 func (p *TextParser) AcceptQuotedString() (string, bool, error) {
 	r, _, err := p.ReadRune()
 	switch {
@@ -33,22 +31,33 @@ func (p *TextParser) AcceptQuotedString() (string, bool, error) {
 }
 
 func lexQuotedString(p *TextParser) (string, *lexer.Error) {
+	s, ok, err := lexQuotedStringNoEscape(p)
+	switch {
+	case err != nil:
+		return "", err
+	case ok:
+		return s, nil
+	default:
+		// escape character detected
+		return lexQuotedStringEscaped(p)
+	}
+}
+
+func lexQuotedStringNoEscape(p *TextParser) (string, bool, *lexer.Error) {
 	for {
 		r, _, err := p.ReadRune()
 		switch {
 		case err != nil:
 			// incomplete
-			return "", NewErrIncompleteQuotedString(p)
+			return "", false, NewErrIncompleteQuotedString(p)
 		case r == RuneQuotes:
-			// end, remove quotes and process escaped characters
-			return lexReturnUnescapedQuotedString(p)
+			// end, just remove the quotes
+			s := p.String()
+			l := len(s)
+			return s[1 : l-2], true, nil
 		case r == RuneEscape:
-			// escaped, take another
-			_, _, err := p.ReadRune()
-			if err != nil {
-				// incomplete
-				return "", NewErrIncompleteQuotedString(p)
-			}
+			// things just got complicated...
+			return "", false, nil
 		case IsNewLine(r):
 			// new lines within quoted values are acceptable
 			p.UnreadRune()
@@ -59,22 +68,11 @@ func lexQuotedString(p *TextParser) (string, *lexer.Error) {
 	}
 }
 
-func lexReturnUnescapedQuotedString(p *TextParser) (string, *lexer.Error) {
-	// remove quotes
-	s := p.String()
-	l := len(s)
-	s = s[1 : l-1]
-
-	if strings.ContainsRune(s, RuneEscape) {
-		// TODO: implement unescaping
-		err := NewError(p.Position(), s, "escaped characters", lexer.ErrNotImplemented)
-		return "", err
-	}
-
-	return s, nil
+func lexQuotedStringEscaped(*TextParser) (string, *lexer.Error) {
+	panic("not implemented")
 }
 
-// Unquoted removes quotes the content and unescapes the content
+// Unquoted removes quotes and unescapes the content
 func Unquoted(s string) (string, error) {
 	var p TextParser
 	if s == "" {