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>

modules/git/batch_reader.go

@@ -2,17 +2,48 @@
 // Use of this source code is governed by a MIT-style
 // license that can be found in the LICENSE file.
 
-// +build !gogit
-
 package git
 
 import (
 	"bufio"
 	"bytes"
+	"io"
 	"math"
 	"strconv"
+	"strings"
 )
 
+// CatFileBatch opens git cat-file --batch in the provided repo and returns a stdin pipe, a stdout reader and cancel function
+func CatFileBatch(repoPath string) (*io.PipeWriter, *bufio.Reader, func()) {
+	// Next feed the commits in order into cat-file --batch, followed by their trees and sub trees as necessary.
+	// so let's create a batch stdin and stdout
+	batchStdinReader, batchStdinWriter := io.Pipe()
+	batchStdoutReader, batchStdoutWriter := io.Pipe()
+	cancel := func() {
+		_ = batchStdinReader.Close()
+		_ = batchStdinWriter.Close()
+		_ = batchStdoutReader.Close()
+		_ = batchStdoutWriter.Close()
+	}
+
+	go func() {
+		stderr := strings.Builder{}
+		err := NewCommand("cat-file", "--batch").RunInDirFullPipeline(repoPath, batchStdoutWriter, &stderr, batchStdinReader)
+		if err != nil {
+			_ = batchStdoutWriter.CloseWithError(ConcatenateError(err, (&stderr).String()))
+			_ = batchStdinReader.CloseWithError(ConcatenateError(err, (&stderr).String()))
+		} else {
+			_ = batchStdoutWriter.Close()
+			_ = batchStdinReader.Close()
+		}
+	}()
+
+	// For simplicities sake we'll us a buffered reader to read from the cat-file --batch
+	batchReader := bufio.NewReader(batchStdoutReader)
+
+	return batchStdinWriter, batchReader, cancel
+}
+
 // ReadBatchLine reads the header line from cat-file --batch
 // We expect:
 // <sha> SP <type> SP <size> LF

modules/git/commit_info_nogogit.go

@@ -141,29 +141,8 @@ func GetLastCommitForPaths(commit *Commit, treePath string, paths []string) ([]*
 		}
 	}()
 
-	// We feed the commits in order into cat-file --batch, followed by their trees and sub trees as necessary.
-	// so let's create a batch stdin and stdout
-	batchStdinReader, batchStdinWriter := io.Pipe()
-	batchStdoutReader, batchStdoutWriter := io.Pipe()
-	defer func() {
-		_ = batchStdinReader.Close()
-		_ = batchStdinWriter.Close()
-		_ = batchStdoutReader.Close()
-		_ = batchStdoutWriter.Close()
-	}()
-
-	go func() {
-		stderr := strings.Builder{}
-		err := NewCommand("cat-file", "--batch").RunInDirFullPipeline(commit.repo.Path, batchStdoutWriter, &stderr, batchStdinReader)
-		if err != nil {
-			_ = revListWriter.CloseWithError(ConcatenateError(err, (&stderr).String()))
-		} else {
-			_ = revListWriter.Close()
-		}
-	}()
-
-	// For simplicities sake we'll us a buffered reader
-	batchReader := bufio.NewReader(batchStdoutReader)
+	batchStdinWriter, batchReader, cancel := CatFileBatch(commit.repo.Path)
+	defer cancel()
 
 	mapsize := 4096
 	if len(paths) > mapsize {

modules/git/pipeline/lfs_nogogit.go

@@ -64,27 +64,8 @@ func FindLFSFile(repo *git.Repository, hash git.SHA1) ([]*LFSResult, error) {
 
 	// Next feed the commits in order into cat-file --batch, followed by their trees and sub trees as necessary.
 	// so let's create a batch stdin and stdout
-	batchStdinReader, batchStdinWriter := io.Pipe()
-	batchStdoutReader, batchStdoutWriter := io.Pipe()
-	defer func() {
-		_ = batchStdinReader.Close()
-		_ = batchStdinWriter.Close()
-		_ = batchStdoutReader.Close()
-		_ = batchStdoutWriter.Close()
-	}()
-
-	go func() {
-		stderr := strings.Builder{}
-		err := git.NewCommand("cat-file", "--batch").RunInDirFullPipeline(repo.Path, batchStdoutWriter, &stderr, batchStdinReader)
-		if err != nil {
-			_ = revListWriter.CloseWithError(git.ConcatenateError(err, (&stderr).String()))
-		} else {
-			_ = revListWriter.Close()
-		}
-	}()
-
-	// For simplicities sake we'll us a buffered reader to read from the cat-file --batch
-	batchReader := bufio.NewReader(batchStdoutReader)
+	batchStdinWriter, batchReader, cancel := git.CatFileBatch(repo.Path)
+	defer cancel()
 
 	// We'll use a scanner for the revList because it's simpler than a bufio.Reader
 	scan := bufio.NewScanner(revListReader)

modules/git/repo_language_stats_nogogit.go

@@ -11,7 +11,6 @@ import (
 	"bytes"
 	"io"
 	"math"
-	"strings"
 
 	"code.gitea.io/gitea/modules/analyze"
 
@@ -22,30 +21,8 @@ import (
 func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, error) {
 	// We will feed the commit IDs in order into cat-file --batch, followed by blobs as necessary.
 	// so let's create a batch stdin and stdout
-
-	batchStdinReader, batchStdinWriter := io.Pipe()
-	batchStdoutReader, batchStdoutWriter := io.Pipe()
-	defer func() {
-		_ = batchStdinReader.Close()
-		_ = batchStdinWriter.Close()
-		_ = batchStdoutReader.Close()
-		_ = batchStdoutWriter.Close()
-	}()
-
-	go func() {
-		stderr := strings.Builder{}
-		err := NewCommand("cat-file", "--batch").RunInDirFullPipeline(repo.Path, batchStdoutWriter, &stderr, batchStdinReader)
-		if err != nil {
-			_ = batchStdoutWriter.CloseWithError(ConcatenateError(err, (&stderr).String()))
-			_ = batchStdinReader.CloseWithError(ConcatenateError(err, (&stderr).String()))
-		} else {
-			_ = batchStdoutWriter.Close()
-			_ = batchStdinReader.Close()
-		}
-	}()
-
-	// For simplicities sake we'll us a buffered reader
-	batchReader := bufio.NewReader(batchStdoutReader)
+	batchStdinWriter, batchReader, cancel := CatFileBatch(repo.Path)
+	defer cancel()
 
 	writeID := func(id string) error {
 		_, err := batchStdinWriter.Write([]byte(id))