Sync branches to DB immediately when handle git hook calling (gitea#29493)

Unlike other async processing in the queue, we should sync branches to
the DB immediately when handling git hook calling. If it fails, users
can see the error message in the output of the git command.

It can avoid potential inconsistency issues, and help #29494.

---------

Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>

services/repository/branch.go

@@ -225,44 +225,91 @@ func checkBranchName(ctx context.Context, repo *repo_model.Repository, name stri
 	return err
 }
 
-// syncBranchToDB sync the branch information in the database. It will try to update the branch first,
-// if updated success with affect records > 0, then all are done. Because that means the branch has been in the database.
-// If no record is affected, that means the branch does not exist in database. So there are two possibilities.
-// One is this is a new branch, then we just need to insert the record. Another is the branches haven't been synced,
-// then we need to sync all the branches into database.
-func syncBranchToDB(ctx context.Context, repoID, pusherID int64, branchName string, commit *git.Commit) error {
-	cnt, err := git_model.UpdateBranch(ctx, repoID, pusherID, branchName, commit)
-	if err != nil {
-		return fmt.Errorf("git_model.UpdateBranch %d:%s failed: %v", repoID, branchName, err)
-	}
-	if cnt > 0 { // This means branch does exist, so it's a normal update. It also means the branch has been synced.
-		return nil
+// SyncBranchesToDB sync the branch information in the database.
+// It will check whether the branches of the repository have never been synced before.
+// If so, it will sync all branches of the repository.
+// Otherwise, it will sync the branches that need to be updated.
+func SyncBranchesToDB(ctx context.Context, repoID, pusherID int64, branchNames, commitIDs []string, getCommit func(commitID string) (*git.Commit, error)) error {
+	// Some designs that make the code look strange but are made for performance optimization purposes:
+	// 1. Sync branches in a batch to reduce the number of DB queries.
+	// 2. Lazy load commit information since it may be not necessary.
+	// 3. Exit early if synced all branches of git repo when there's no branch in DB.
+	// 4. Check the branches in DB if they are already synced.
+	//
+	// If the user pushes many branches at once, the Git hook will call the internal API in batches, rather than all at once.
+	// See https://github.com/go-gitea/gitea/blob/cb52b17f92e2d2293f7c003649743464492bca48/cmd/hook.go#L27
+	// For the first batch, it will hit optimization 3.
+	// For other batches, it will hit optimization 4.
+
+	if len(branchNames) != len(commitIDs) {
+		return fmt.Errorf("branchNames and commitIDs length not match")
 	}
 
-	// if user haven't visit UI but directly push to a branch after upgrading from 1.20 -> 1.21,
-	// we cannot simply insert the branch but need to check we have branches or not
-	hasBranch, err := db.Exist[git_model.Branch](ctx, git_model.FindBranchOptions{
-		RepoID:          repoID,
-		IsDeletedBranch: optional.Some(false),
-	}.ToConds())
-	if err != nil {
-		return err
-	}
-	if !hasBranch {
-		if _, err = repo_module.SyncRepoBranches(ctx, repoID, pusherID); err != nil {
-			return fmt.Errorf("repo_module.SyncRepoBranches %d:%s failed: %v", repoID, branchName, err)
+	return db.WithTx(ctx, func(ctx context.Context) error {
+		branches, err := git_model.GetBranches(ctx, repoID, branchNames)
+		if err != nil {
+			return fmt.Errorf("git_model.GetBranches: %v", err)
+		}
+
+		if len(branches) == 0 {
+			// if user haven't visit UI but directly push to a branch after upgrading from 1.20 -> 1.21,
+			// we cannot simply insert the branch but need to check we have branches or not
+			hasBranch, err := db.Exist[git_model.Branch](ctx, git_model.FindBranchOptions{
+				RepoID:          repoID,
+				IsDeletedBranch: optional.Some(false),
+			}.ToConds())
+			if err != nil {
+				return err
+			}
+			if !hasBranch {
+				if _, err = repo_module.SyncRepoBranches(ctx, repoID, pusherID); err != nil {
+					return fmt.Errorf("repo_module.SyncRepoBranches %d failed: %v", repoID, err)
+				}
+				return nil
+			}
+		}
+
+		branchMap := make(map[string]*git_model.Branch, len(branches))
+		for _, branch := range branches {
+			branchMap[branch.Name] = branch
+		}
+
+		newBranches := make([]*git_model.Branch, 0, len(branchNames))
+
+		for i, branchName := range branchNames {
+			commitID := commitIDs[i]
+			branch, exist := branchMap[branchName]
+			if exist && branch.CommitID == commitID {
+				continue
+			}
+
+			commit, err := getCommit(branchName)
+			if err != nil {
+				return fmt.Errorf("get commit of %s failed: %v", branchName, err)
+			}
+
+			if exist {
+				if _, err := git_model.UpdateBranch(ctx, repoID, pusherID, branchName, commit); err != nil {
+					return fmt.Errorf("git_model.UpdateBranch %d:%s failed: %v", repoID, branchName, err)
+				}
+				return nil
+			}
+
+			// if database have branches but not this branch, it means this is a new branch
+			newBranches = append(newBranches, &git_model.Branch{
+				RepoID:        repoID,
+				Name:          branchName,
+				CommitID:      commit.ID.String(),
+				CommitMessage: commit.Summary(),
+				PusherID:      pusherID,
+				CommitTime:    timeutil.TimeStamp(commit.Committer.When.Unix()),
+			})
+		}
+
+		if len(newBranches) > 0 {
+			return db.Insert(ctx, newBranches)
 		}
 		return nil
-	}
-
-	// if database have branches but not this branch, it means this is a new branch
-	return db.Insert(ctx, &git_model.Branch{
-		RepoID:        repoID,
-		Name:          branchName,
-		CommitID:      commit.ID.String(),
-		CommitMessage: commit.Summary(),
-		PusherID:      pusherID,
-		CommitTime:    timeutil.TimeStamp(commit.Committer.When.Unix()),
 	})
 }
 

services/repository/push.go

@@ -11,7 +11,6 @@ import (
 	"time"
 
 	"code.gitea.io/gitea/models/db"
-	git_model "code.gitea.io/gitea/models/git"
 	repo_model "code.gitea.io/gitea/models/repo"
 	user_model "code.gitea.io/gitea/models/user"
 	"code.gitea.io/gitea/modules/cache"
@@ -259,10 +258,6 @@ func pushUpdates(optsList []*repo_module.PushUpdateOptions) error {
 					commits.Commits = commits.Commits[:setting.UI.FeedMaxCommitNum]
 				}
 
-				if err = syncBranchToDB(ctx, repo.ID, opts.PusherID, branch, newCommit); err != nil {
-					return fmt.Errorf("git_model.UpdateBranch %s:%s failed: %v", repo.FullName(), branch, err)
-				}
-
 				notify_service.PushCommits(ctx, pusher, repo, opts, commits)
 
 				// Cache for big repository
@@ -275,10 +270,6 @@ func pushUpdates(optsList []*repo_module.PushUpdateOptions) error {
 					// close all related pulls
 					log.Error("close related pull request failed: %v", err)
 				}
-
-				if err := git_model.AddDeletedBranch(ctx, repo.ID, branch, pusher.ID); err != nil {
-					return fmt.Errorf("AddDeletedBranch %s:%s failed: %v", repo.FullName(), branch, err)
-				}
 			}
 
 			// Even if user delete a branch on a repository which he didn't watch, he will be watch that.