Fix Parser.Range returning wrong offset for error highlights

The subsliceOffset method incorrectly computed offset as
len(p.data) - len(b), which only works when b is a suffix (tail) of
p.data. However, error highlights (ParserError.Highlight) are arbitrary
subslices from the middle of the input (e.g., b[0:1] from parseSimpleKey),
so their len has no relationship to their position.

This was a regression introduced in commit 3aaf147 (Remove unsafe package
usage) which replaced danger.SubsliceOffset (pointer arithmetic) with the
incorrect len-based approach.

Fix by using reflect.ValueOf().Pointer() to compute the actual byte
offset between slice data pointers, matching the approach already used
in errors.go:subsliceOffset.

Fixes #1047

Co-authored-by: Thomas Pelletier <thomas@pelletier.dev>

unstable/parser.go

@@ -3,6 +3,7 @@ package unstable
 import (
 	"bytes"
 	"fmt"
+	"reflect"
 	"unicode"
 
 	"github.com/pelletier/go-toml/v2/internal/characters"
@@ -83,10 +84,18 @@ func (p *Parser) rangeOfToken(token, rest []byte) Range {
 }
 
 // subsliceOffset returns the byte offset of subslice b within p.data.
-// b must be a suffix (tail) of p.data.
+// b must share the same backing array as p.data.
 func (p *Parser) subsliceOffset(b []byte) int {
-	// b is a suffix of p.data, so its offset is len(p.data) - len(b)
-	return len(p.data) - len(b)
+	if len(b) == 0 {
+		return 0
+	}
+	dataPtr := reflect.ValueOf(p.data).Pointer()
+	subPtr := reflect.ValueOf(b).Pointer()
+	offset := int(subPtr - dataPtr)
+	if offset < 0 || offset > len(p.data) {
+		panic("subslice is not within parser data")
+	}
+	return offset
 }
 
 // Raw returns the slice corresponding to the bytes in the given range.