(incubator-devlake) branch release-v1.0 updated: fix(gitextractor): support empty repos (#7963) (#7964)

[ly...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

lynwee pushed a commit to branch release-v1.0
in repository https://gitbox.apache.org/repos/asf/incubator-devlake.git


The following commit(s) were added to refs/heads/release-v1.0 by this push:
     new cc7250067 fix(gitextractor): support empty repos (#7963) (#7964)
cc7250067 is described below

commit cc7250067c714b266b231c77f5d6bea65c4b5d52
Author: github-actions[bot] <41...@users.noreply.github.com>
AuthorDate: Tue Aug 27 16:30:43 2024 +0800

    fix(gitextractor): support empty repos (#7963) (#7964)
    
    * fix(gitextractor): support empty repos
    
    * fix(gitextractor): tiny fix
    
    Co-authored-by: Lynwee <15...@qq.com>
---
 .../plugins/gitextractor/parser/clone_gitcli.go    | 35 ++++++++++++++++++++--
 1 file changed, 33 insertions(+), 2 deletions(-)

diff --git a/backend/plugins/gitextractor/parser/clone_gitcli.go b/backend/plugins/gitextractor/parser/clone_gitcli.go
index 6dbf5481a..adeceb9e3 100644
--- a/backend/plugins/gitextractor/parser/clone_gitcli.go
+++ b/backend/plugins/gitextractor/parser/clone_gitcli.go
@@ -258,7 +258,7 @@ func (g *GitcliCloner) doubleClone() errors.Error {
 		return err
 	}
 	g.localDir = backup
-	// step 2: perform shallow clone aginst the intermediary dir
+	// step 2: perform shallow clone against the intermediary dir
 	backup = g.remoteUrl
 	g.remoteUrl = fmt.Sprintf("file://%s", intermediaryDir) // the file:// prefix is required for shallow clone to work
 	if err := g.shallowClone(); err != nil {
@@ -274,10 +274,30 @@ func (g *GitcliCloner) gitClone(args ...string) errors.Error {
 }
 
 func (g *GitcliCloner) gitFetch(args ...string) errors.Error {
+	empty, err := g.repoIsEmpty(args...)
+	if err != nil {
+		g.logger.Error(err, "repo is empty")
+		return err
+	}
+	if empty {
+		g.logger.Info("repo is empty, doesn't need to fetch")
+		return nil
+	}
 	args = append(args, g.syncArgs...)
 	return g.git(g.syncEnvs, g.localDir, "fetch", args...)
 }
 
+func (g *GitcliCloner) repoIsEmpty(args ...string) (bool, errors.Error) {
+	// try to run command: git log
+	// if repo is empty, it will return an error
+	err := g.git(g.syncEnvs, g.localDir, "log")
+	if err != nil {
+		g.logger.Warn(err, "git log failed")
+		return true, nil
+	}
+	return false, nil
+}
+
 func (g *GitcliCloner) gitCmd(gitcmd string, args ...string) errors.Error {
 	return g.git(nil, g.localDir, gitcmd, args...)
 }
@@ -294,15 +314,26 @@ func (g *GitcliCloner) git(env []string, dir string, gitcmd string, args ...stri
 func (g *GitcliCloner) execCommand(cmd *exec.Cmd) errors.Error {
 	output, err := cmd.CombinedOutput()
 	if err != nil {
+		g.logger.Debug("err: %v, output: %s", err, string(output))
 		outputString := string(output)
 		if strings.Contains(outputString, "fatal: error processing shallow info: 4") ||
 			strings.Contains(outputString, "fatal: the remote end hung up unexpectedly") {
 			return ErrNoData
 		}
-		return errors.Default.New(fmt.Sprintf("git cmd %v in %s failed: %s", sanitizeArgs(cmd.Args), cmd.Dir, outputString))
+		return errors.Default.New(fmt.Sprintf("git cmd %v in %s failed: %s", sanitizeArgs(cmd.Args), cmd.Dir, generateErrMsg(output, err)))
 	}
 	return nil
 }
+func generateErrMsg(output []byte, err error) string {
+	errMsg := strings.TrimSpace(string(output))
+	if errMsg == "" {
+		errMsg = err.Error()
+	}
+	if errMsg == "" {
+		errMsg = "unknown error"
+	}
+	return errMsg
+}
 
 func sanitizeArgs(args []string) []string {
 	var ret []string