Merge pull request weaviate#7837 from weaviate/optimize_header_parsing

Optimize Segmentindex header parsing

Author: dirkkul

adapters/repos/db/lsmkv/segmentindex/header.go

@@ -17,6 +17,8 @@ import (
 	"encoding/binary"
 	"fmt"
 	"io"
+
+	"github.com/weaviate/weaviate/usecases/byteops"
 )
 
 const (
@@ -40,20 +42,20 @@ type Header struct {
 }
 
 func (h *Header) WriteTo(w io.Writer) (int64, error) {
-	if err := binary.Write(w, binary.LittleEndian, &h.Level); err != nil {
-		return -1, err
-	}
-	if err := binary.Write(w, binary.LittleEndian, &h.Version); err != nil {
-		return -1, err
-	}
-	if err := binary.Write(w, binary.LittleEndian, &h.SecondaryIndices); err != nil {
-		return -1, err
-	}
-	if err := binary.Write(w, binary.LittleEndian, h.Strategy); err != nil {
-		return -1, err
+	data := make([]byte, HeaderSize)
+	rw := byteops.NewReadWriter(data)
+	rw.WriteUint16(h.Level)
+	rw.WriteUint16(h.Version)
+	rw.WriteUint16(h.SecondaryIndices)
+	rw.WriteUint16(uint16(h.Strategy))
+	rw.WriteUint64(h.IndexStart)
+
+	write, err := w.Write(data)
+	if err != nil {
+		return 0, err
 	}
-	if err := binary.Write(w, binary.LittleEndian, &h.IndexStart); err != nil {
-		return -1, err
+	if write != HeaderSize {
+		return 0, fmt.Errorf("expected to write %d bytes, got %d", HeaderSize, write)
 	}
 
 	return int64(HeaderSize), nil
@@ -113,31 +115,26 @@ func (h *Header) SecondaryIndex(source []byte, indexID uint16) ([]byte, error) {
 }
 
 func ParseHeader(r io.Reader) (*Header, error) {
-	out := &Header{}
+	data := make([]byte, HeaderSize)
 
-	if err := binary.Read(r, binary.LittleEndian, &out.Level); err != nil {
+	full, err := io.ReadFull(r, data)
+	if err != nil {
 		return nil, err
 	}
-
-	if err := binary.Read(r, binary.LittleEndian, &out.Version); err != nil {
-		return nil, err
-	}
-
-	if err := binary.Read(r, binary.LittleEndian, &out.SecondaryIndices); err != nil {
-		return nil, err
+	if full != HeaderSize {
+		return nil, fmt.Errorf("expected %d bytes, got %d", HeaderSize, full)
 	}
+	rw := byteops.NewReadWriter(data)
+	out := &Header{}
+	out.Level = rw.ReadUint16()
+	out.Version = rw.ReadUint16()
+	out.SecondaryIndices = rw.ReadUint16()
+	out.Strategy = Strategy(rw.ReadUint16())
+	out.IndexStart = rw.ReadUint64()
 
 	if out.Version > CurrentSegmentVersion {
 		return nil, fmt.Errorf("unsupported version %d", out.Version)
 	}
 
-	if err := binary.Read(r, binary.LittleEndian, &out.Strategy); err != nil {
-		return nil, err
-	}
-
-	if err := binary.Read(r, binary.LittleEndian, &out.IndexStart); err != nil {
-		return nil, err
-	}
-
 	return out, nil
 }

adapters/repos/db/lsmkv/segmentindex/header_test.go

@@ -0,0 +1,47 @@
+//                           _       _
+// __      _____  __ ___   ___  __ _| |_ ___
+// \ \ /\ / / _ \/ _` \ \ / / |/ _` | __/ _ \
+//  \ V  V /  __/ (_| |\ V /| | (_| | ||  __/
+//   \_/\_/ \___|\__,_| \_/ |_|\__,_|\__\___|
+//
+//  Copyright © 2016 - 2024 Weaviate B.V. All rights reserved.
+//
+//  CONTACT: [email protected]
+//
+
+package segmentindex
+
+import (
+	"bytes"
+	"os"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+)
+
+func BenchmarkParseHeader(b *testing.B) {
+	data := []byte{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
+	require.Len(b, data, HeaderSize)
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		_, err := ParseHeader(bytes.NewReader(data))
+		require.NoError(b, err)
+	}
+}
+
+func BenchmarkWriteHeader(b *testing.B) {
+	header := Header{
+		Version:          1,
+		Level:            1,
+		SecondaryIndices: 35,
+		Strategy:         StrategyReplace,
+		IndexStart:       234,
+	}
+	path := b.TempDir()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		f, err := os.Create(path + "/test.tmp")
+		require.NoError(b, err)
+		header.WriteTo(f)
+	}
+}