Merge pull request #116 from devtron-labs/git-cli-analytics-develop

fix: Git cli analytics develop

Author: ayu-devtron

config.md

@@ -1,16 +1,18 @@
 # GITSENSOR CONFIGMAP
 
-| Key                               | Value                              | Description                   |
-|-----------------------------------|------------------------------------|-------------------------------|
-| PG_ADDR                           | postgresql-postgresql.devtroncd   | PostgreSQL Server Address      |
-| PG_USER                           | postgres                           | PostgreSQL User               |
-| PG_DATABASE                       | git_sensor                          | PostgreSQL Database Name       |
-| SENTRY_ENV                        | prod                           | Sentry Environment             |
-| SENTRY_ENABLED                    | "false"                            | Sentry Enabled (boolean)      |
-| POLL_DURATION                     | "1"                                | Polling Duration (in seconds)  |
-| POLL_WORKER                       | "2"                                | Number of Polling Workers     |
-| PG_LOG_QUERY                      | "false"                            | PostgreSQL Query Logging (boolean) |
-| COMMIT_STATS_TIMEOUT_IN_SEC       | "2"                                | Commit Stats Timeout (in seconds) |
-| ENABLE_FILE_STATS                | "false"                            | Enable File Stats (boolean)   |
-| GIT_HISTORY_COUNT                 | "15"                               | Git History Count              |
-| CLONING_MODE                     | FULL                               | Cloning Mode (Possible values: SHALLOW, FULL)       |
+| Key                         | Value                           | Description                                                         |
+|-----------------------------|---------------------------------|---------------------------------------------------------------------|
+| PG_ADDR                     | postgresql-postgresql.devtroncd | PostgreSQL Server Address                                           |
+| PG_USER                     | postgres                        | PostgreSQL User                                                     |
+| PG_DATABASE                 | git_sensor                      | PostgreSQL Database Name                                            |
+| SENTRY_ENV                  | prod                            | Sentry Environment                                                  |
+| SENTRY_ENABLED              | "false"                         | Sentry Enabled (boolean)                                            |
+| POLL_DURATION               | "1"                             | Polling Duration (in seconds)                                       |
+| POLL_WORKER                 | "2"                             | Number of Polling Workers                                           |
+| PG_LOG_QUERY                | "false"                         | PostgreSQL Query Logging (boolean)                                  |
+| COMMIT_STATS_TIMEOUT_IN_SEC | "2"                             | Commit Stats Timeout (in seconds)                                   |
+| ENABLE_FILE_STATS           | "false"                         | Enable File Stats (boolean)                                         |
+| GIT_HISTORY_COUNT           | "15"                            | Git History Count                                                   |
+| CLONING_MODE                | FULL                            | Cloning Mode (Possible values: SHALLOW, FULL)                       |
+| USE_GIT_CLI                 | "false"                         | Use git cli commands directly for all git operations                |
+| USE_GIT_CLI_ANALYTICS       | "false"                         | Use git cli commands directly for getting commit data for analytics |

internals/Configuration.go

@@ -24,10 +24,11 @@ type Configuration struct {
 	GitHistoryCount         int    `env:"GIT_HISTORY_COUNT" envDefault:"15"`
 	MinLimit                int    `env:"MIN_LIMIT_FOR_PVC" envDefault:"1"` // in MB
 	UseGitCli               bool   `env:"USE_GIT_CLI" envDefault:"false"`
+	UseGitCliAnalytics      bool   `env:"USE_GIT_CLI_ANALYTICS" envDefault:"false"` // This flag is used to compute commitDiff using git-cli only for analytics
 	AnalyticsDebug          bool   `env:"ANALYTICS_DEBUG" envDefault:"false"`
 	CliCmdTimeoutGlobal     int    `env:"CLI_CMD_TIMEOUT_GLOBAL_SECONDS" envDefault:"0"`
 	CliCmdTimeoutJson       string `env:"CLI_CMD_TIMEOUT_JSON" envDefault:""`
-	GoGitTimeout            int    `env:"GOGIT_TIMEOUT_SECONDS" envDefault:"10"`
+	GoGitTimeout            int    `env:"GOGIT_TIMEOUT_SECONDS" envDefault:"10" `
 }
 
 func ParseConfiguration() (*Configuration, error) {

pkg/git/RepositoryManagerAnalytics.go

@@ -202,6 +202,7 @@ func transform(src *object.Commit, tag *object.Tag) (dst *Commit) {
 func (impl RepositoryManagerAnalyticsImpl) ChangesSinceByRepositoryForAnalytics(gitCtx GitContext, checkoutPath string, Old string, New string) (*GitChanges, error) {
 	var err error
 	start := time.Now()
+	useGitCli := impl.configuration.UseGitCli || impl.configuration.UseGitCliAnalytics
 	defer func() {
 		util.TriggerGitOperationMetrics("changesSinceByRepositoryForAnalytics", start, err)
 	}()
@@ -214,7 +215,7 @@ func (impl RepositoryManagerAnalyticsImpl) ChangesSinceByRepositoryForAnalytics(
 	oldHash := plumbing.NewHash(Old)
 
 	var fileStats FileStats
-	if strings.Contains(checkoutPath, "/.git") || impl.configuration.UseGitCli {
+	if strings.Contains(checkoutPath, "/.git") || useGitCli {
 		oldHashString := oldHash.String()
 		newHashString := newHash.String()
 		fileStats, err = impl.gitManager.FetchDiffStatBetweenCommitsWithNumstat(gitCtx, oldHashString, newHashString, checkoutPath)
@@ -243,14 +244,17 @@ func (impl RepositoryManagerAnalyticsImpl) ChangesSinceByRepositoryForAnalytics(
 func (impl RepositoryManagerAnalyticsImpl) computeCommitDiff(gitCtx GitContext, checkoutPath string, oldHash plumbing.Hash, newHash plumbing.Hash, repository *GitRepository) ([]*Commit, error) {
 	var commitsCli, commitsGoGit []*Commit
 	var err error
-	if impl.configuration.UseGitCli || impl.configuration.AnalyticsDebug {
+	useGitCli := impl.configuration.UseGitCli || impl.configuration.UseGitCliAnalytics
+	if useGitCli || impl.configuration.AnalyticsDebug {
+		impl.logger.Infow("Computing commit diff using cli ", "checkoutPath", checkoutPath)
 		commitsCli, err = impl.gitManager.LogMergeBase(gitCtx, checkoutPath, oldHash.String(), newHash.String())
 		if err != nil {
 			impl.logger.Errorw("error in fetching commits for analytics through CLI: ", "err", err)
 			return nil, err
 		}
 	}
-	if !impl.configuration.UseGitCli || impl.configuration.AnalyticsDebug {
+	if !useGitCli || impl.configuration.AnalyticsDebug {
+		impl.logger.Infow("Computing commit diff using go-git ", "checkoutPath", checkoutPath)
 		ctx, cancel := gitCtx.WithTimeout(impl.configuration.GoGitTimeout)
 		defer cancel()
 		commitsGoGit, err = RunWithTimeout(ctx, func() ([]*Commit, error) {
@@ -265,7 +269,7 @@ func (impl RepositoryManagerAnalyticsImpl) computeCommitDiff(gitCtx GitContext,
 		impl.logOldestCommitComparison(commitsGoGit, commitsCli, checkoutPath, oldHash.String(), newHash.String())
 	}
 
-	if !impl.configuration.UseGitCli {
+	if !useGitCli {
 		return commitsGoGit, nil
 	}
 	return commitsCli, nil