Merge pull request '[BUG] Handle bigger files in `git grep`' (#3160) from gusted/forgejo-grep into forgejo

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/3160
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Reviewed-by: 0ko <0ko@noreply.codeberg.org>
Reviewed-by: Shiny Nematoda <snematoda@noreply.codeberg.org>

modules/git/grep.go

@@ -10,6 +10,7 @@ import (
 	"context"
 	"errors"
 	"fmt"
+	"io"
 	"os"
 	"strconv"
 	"strings"
@@ -80,10 +81,21 @@ func GrepSearch(ctx context.Context, repo *Repository, search string, opts GrepO
 			defer stdoutReader.Close()
 
 			isInBlock := false
-			scanner := bufio.NewScanner(stdoutReader)
+			scanner := bufio.NewReader(stdoutReader)
 			var res *GrepResult
-			for scanner.Scan() {
+			for {
-				line := scanner.Text()
+				line, err := scanner.ReadString('\n')
+				if err != nil {
+					if err == io.EOF {
+						return nil
+					}
+					return err
+				}
+				// Remove delimiter.
+				if len(line) > 0 {
+					line = line[:len(line)-1]
+				}
+
 				if !isInBlock {
 					if _ /* ref */, filename, ok := strings.Cut(line, ":"); ok {
 						isInBlock = true
@@ -109,7 +121,7 @@ func GrepSearch(ctx context.Context, repo *Repository, search string, opts GrepO
 					res.LineCodes = append(res.LineCodes, lineCode)
 				}
 			}
-			return scanner.Err()
+			return nil
 		},
 	})
 	// git grep exits by cancel (killed), usually it is caused by the limit of results

modules/git/grep_test.go

@@ -4,7 +4,10 @@
 package git
 
 import (
+	"bytes"
 	"context"
+	"os"
+	"path"
 	"path/filepath"
 	"testing"
 
@@ -49,3 +52,27 @@ func TestGrepSearch(t *testing.T) {
 	assert.Error(t, err)
 	assert.Len(t, res, 0)
 }
+
+func TestGrepLongFiles(t *testing.T) {
+	tmpDir := t.TempDir()
+
+	err := InitRepository(DefaultContext, tmpDir, false, Sha1ObjectFormat.Name())
+	assert.NoError(t, err)
+
+	gitRepo, err := openRepositoryWithDefaultContext(tmpDir)
+	assert.NoError(t, err)
+	defer gitRepo.Close()
+
+	assert.NoError(t, os.WriteFile(path.Join(tmpDir, "README.md"), bytes.Repeat([]byte{'a'}, 65*1024), 0o666))
+
+	err = AddChanges(tmpDir, true)
+	assert.NoError(t, err)
+
+	err = CommitChanges(tmpDir, CommitChangesOptions{Message: "Long file"})
+	assert.NoError(t, err)
+
+	res, err := GrepSearch(context.Background(), gitRepo, "a", GrepOptions{})
+	assert.NoError(t, err)
+	assert.Len(t, res, 1)
+	assert.Len(t, res[0].LineCodes[0], 65*1024)
+}