Move Bleve and Elastic code indexers to use a common cat-file --batch (#14781)

* Extract out the common cat-file batch calls

Signed-off-by: Andrew Thornton <art27@cantab.net>

* Move bleve and elastic indexers to use a common cat-file --batch when indexing

Signed-off-by: Andrew Thornton <art27@cantab.net>

* move catfilebatch to batch_reader and rename to batch_reader.go

Signed-off-by: Andrew Thornton <art27@cantab.net>

Co-authored-by: 6543 <6543@obermui.de>
Co-authored-by: Lauris BH <lauris@nix.lv>

2 files changed, 52 insertions, 16 deletions

diff --git a/modules/indexer/code/bleve.go b/modules/indexer/code/bleve.go
index 1ebc74c43a..573ea8b88c 100644
--- a/modules/indexer/code/bleve.go
+++ b/modules/indexer/code/bleve.go
@@ -5,7 +5,10 @@
 package code
 
 import (
+	"bufio"
 	"fmt"
+	"io"
+	"io/ioutil"
 	"os"
 	"strconv"
 	"strings"
@@ -173,7 +176,7 @@ func NewBleveIndexer(indexDir string) (*BleveIndexer, bool, error) {
 	return indexer, created, err
 }
 
-func (b *BleveIndexer) addUpdate(commitSha string, update fileUpdate, repo *models.Repository, batch rupture.FlushingBatch) error {
+func (b *BleveIndexer) addUpdate(batchWriter *io.PipeWriter, batchReader *bufio.Reader, commitSha string, update fileUpdate, repo *models.Repository, batch rupture.FlushingBatch) error {
 	// Ignore vendored files in code search
 	if setting.Indexer.ExcludeVendored && enry.IsVendor(update.Filename) {
 		return nil
@@ -196,8 +199,16 @@ func (b *BleveIndexer) addUpdate(commitSha string, update fileUpdate, repo *mode
 		return b.addDelete(update.Filename, repo, batch)
 	}
 
-	fileContents, err := git.NewCommand("cat-file", "blob", update.BlobSha).
-		RunInDirBytes(repo.RepoPath())
+	if _, err := batchWriter.Write([]byte(update.BlobSha + "\n")); err != nil {
+		return err
+	}
+
+	_, _, size, err := git.ReadBatchLine(batchReader)
+	if err != nil {
+		return err
+	}
+
+	fileContents, err := ioutil.ReadAll(io.LimitReader(batchReader, size))
 	if err != nil {
 		return err
 	} else if !base.IsTextFile(fileContents) {
@@ -254,10 +265,17 @@ func (b *BleveIndexer) Close() {
 // Index indexes the data
 func (b *BleveIndexer) Index(repo *models.Repository, sha string, changes *repoChanges) error {
 	batch := rupture.NewFlushingBatch(b.indexer, maxBatchSize)
-	for _, update := range changes.Updates {
-		if err := b.addUpdate(sha, update, repo, batch); err != nil {
-			return err
+	if len(changes.Updates) > 0 {
+
+		batchWriter, batchReader, cancel := git.CatFileBatch(repo.RepoPath())
+		defer cancel()
+
+		for _, update := range changes.Updates {
+			if err := b.addUpdate(batchWriter, batchReader, sha, update, repo, batch); err != nil {
+				return err
+			}
 		}
+		cancel()
 	}
 	for _, filename := range changes.RemovedFilenames {
 		if err := b.addDelete(filename, repo, batch); err != nil {
diff --git a/modules/indexer/code/elastic_search.go b/modules/indexer/code/elastic_search.go
index 130cd1430a..5327eb1e51 100644
--- a/modules/indexer/code/elastic_search.go
+++ b/modules/indexer/code/elastic_search.go
@@ -5,8 +5,11 @@
 package code
 
 import (
+	"bufio"
 	"context"
 	"fmt"
+	"io"
+	"io/ioutil"
 	"strconv"
 	"strings"
 	"time"
@@ -172,7 +175,7 @@ func (b *ElasticSearchIndexer) init() (bool, error) {
 	return exists, nil
 }
 
-func (b *ElasticSearchIndexer) addUpdate(sha string, update fileUpdate, repo *models.Repository) ([]elastic.BulkableRequest, error) {
+func (b *ElasticSearchIndexer) addUpdate(batchWriter *io.PipeWriter, batchReader *bufio.Reader, sha string, update fileUpdate, repo *models.Repository) ([]elastic.BulkableRequest, error) {
 	// Ignore vendored files in code search
 	if setting.Indexer.ExcludeVendored && enry.IsVendor(update.Filename) {
 		return nil, nil
@@ -195,8 +198,16 @@ func (b *ElasticSearchIndexer) addUpdate(sha string, update fileUpdate, repo *mo
 		return []elastic.BulkableRequest{b.addDelete(update.Filename, repo)}, nil
 	}
 
-	fileContents, err := git.NewCommand("cat-file", "blob", update.BlobSha).
-		RunInDirBytes(repo.RepoPath())
+	if _, err := batchWriter.Write([]byte(update.BlobSha + "\n")); err != nil {
+		return nil, err
+	}
+
+	_, _, size, err := git.ReadBatchLine(batchReader)
+	if err != nil {
+		return nil, err
+	}
+
+	fileContents, err := ioutil.ReadAll(io.LimitReader(batchReader, size))
 	if err != nil {
 		return nil, err
 	} else if !base.IsTextFile(fileContents) {
@@ -230,14 +241,21 @@ func (b *ElasticSearchIndexer) addDelete(filename string, repo *models.Repositor
 // Index will save the index data
 func (b *ElasticSearchIndexer) Index(repo *models.Repository, sha string, changes *repoChanges) error {
 	reqs := make([]elastic.BulkableRequest, 0)
-	for _, update := range changes.Updates {
-		updateReqs, err := b.addUpdate(sha, update, repo)
-		if err != nil {
-			return err
-		}
-		if len(updateReqs) > 0 {
-			reqs = append(reqs, updateReqs...)
+	if len(changes.Updates) > 0 {
+
+		batchWriter, batchReader, cancel := git.CatFileBatch(repo.RepoPath())
+		defer cancel()
+
+		for _, update := range changes.Updates {
+			updateReqs, err := b.addUpdate(batchWriter, batchReader, sha, update, repo)
+			if err != nil {
+				return err
+			}
+			if len(updateReqs) > 0 {
+				reqs = append(reqs, updateReqs...)
+			}
 		}
+		cancel()
 	}
 
 	for _, filename := range changes.RemovedFilenames {