📝 bytestrings: add iterator over non-empty lines

bytestrings/bytestrings.go

@@ -1,11 +1,13 @@
 package bytestrings
 
 import (
+	"iter"
 	"strings"
 	"unsafe"
 )
 
 // NextNonEmptyLine returns the next non-empty line and the remaining text.
+// The line has its line feed ('\n') and carriage return ('\r') characters removed.
 func NextNonEmptyLine[T ~[]byte | ~string](text T) (T, T) {
 	for {
 		lfIndex := strings.IndexByte(*(*string)(unsafe.Pointer(&text)), '\n')
@@ -26,3 +28,31 @@ func NextNonEmptyLine[T ~[]byte | ~string](text T) (T, T) {
 		return line, text
 	}
 }
+
+// NonEmptyLines returns an iterator over the non-empty lines in the text,
+// with line feed ('\n') and carriage return ('\r') characters removed.
+func NonEmptyLines[T ~[]byte | ~string](text T) iter.Seq[T] {
+	return func(yield func(T) bool) {
+		for lfIndex := 0; len(text) > 0; text = text[lfIndex+1:] {
+			lfIndex = strings.IndexByte(*(*string)(unsafe.Pointer(&text)), '\n')
+			switch lfIndex {
+			case -1:
+				_ = yield(text)
+				return
+			case 0:
+				continue
+			}
+
+			line := text[:lfIndex]
+			if line[len(line)-1] == '\r' {
+				line = line[:len(line)-1]
+			}
+			if len(line) == 0 {
+				continue
+			}
+			if !yield(line) {
+				return
+			}
+		}
+	}
+}

bytestrings/bytestrings_test.go

@@ -1,6 +1,9 @@
 package bytestrings
 
-import "testing"
+import (
+	"slices"
+	"testing"
+)
 
 const multiline = "\n1\r\n2\n\n3\r\n\r\n4"
 
@@ -37,3 +40,11 @@ func TestNextNonEmptyLine(t *testing.T) {
 		t.Fatalf("Expected text '%s', got '%s'", multiline[13:], text)
 	}
 }
+
+func TestNonEmptyLines(t *testing.T) {
+	expectedLines := []string{"1", "2", "3", "4"}
+	lines := slices.AppendSeq(make([]string, 0, len(expectedLines)), NonEmptyLines(multiline))
+	if !slices.Equal(lines, expectedLines) {
+		t.Errorf("Expected lines %v, got %v", expectedLines, lines)
+	}
+}

cmd/shadowsocks-go-domain-set-converter/main.go

@@ -128,14 +128,7 @@ func DomainSetBuilderFromDlc(text string) (domainset.Builder, error) {
 		domainset.NewRegexpMatcherBuilder(0),
 	}
 
-	var line string
-
-	for {
-		line, text = bytestrings.NextNonEmptyLine(text)
-		if len(line) == 0 {
-			break
-		}
-
+	for line := range bytestrings.NonEmptyLines(text) {
 		if line[0] == '#' {
 			continue
 		}

domainset/domainset.go

@@ -6,6 +6,7 @@ import (
 	"errors"
 	"fmt"
 	"io"
+	"iter"
 	"strconv"
 	"strings"
 
@@ -206,8 +207,11 @@ func BuilderFromTextFunc(
 	newKeywordMatcherBuilderFunc,
 	newRegexpMatcherBuilderFunc func(int) MatcherBuilder,
 ) (Builder, error) {
-	line, text := bytestrings.NextNonEmptyLine(text)
-	if len(line) == 0 {
+	next, stop := iter.Pull(bytestrings.NonEmptyLines(text))
+	defer stop()
+
+	line, ok := next()
+	if !ok {
 		return Builder{}, errEmptySet
 	}
 
@@ -216,8 +220,8 @@ func BuilderFromTextFunc(
 		return Builder{}, err
 	}
 	if found {
-		line, text = bytestrings.NextNonEmptyLine(text)
-		if len(line) == 0 {
+		line, ok = next()
+		if !ok {
 			return Builder{}, errEmptySet
 		}
 	}
@@ -255,8 +259,8 @@ func BuilderFromTextFunc(
 		}
 
 	next:
-		line, text = bytestrings.NextNonEmptyLine(text)
-		if len(line) == 0 {
+		line, ok = next()
+		if !ok {
 			break
 		}
 	}

prefixset/prefixset.go

@@ -28,17 +28,9 @@ func (psc Config) IPSet() (*netipx.IPSet, error) {
 
 // IPSetFromText parses prefixes from the text and builds a prefix set.
 func IPSetFromText(text string) (*netipx.IPSet, error) {
-	var (
-		line string
-		sb   netipx.IPSetBuilder
-	)
-
-	for {
-		line, text = bytestrings.NextNonEmptyLine(text)
-		if len(line) == 0 {
-			break
-		}
+	var sb netipx.IPSetBuilder
 
+	for line := range bytestrings.NonEmptyLines(text) {
 		if line[0] == '#' {
 			continue
 		}