Change language statistics to save size instead of percentage (#11681)

* Change language statistics to save size instead of percentage in database

Co-Authored-By: Cirno the Strongest <1447794+CirnoT@users.noreply.github.com>

* Do not exclude if only language

* Fix edge cases with special langauges

Co-authored-by: Cirno the Strongest <1447794+CirnoT@users.noreply.github.com>

2 files changed, 9 insertions, 19 deletions

diff --git a/modules/git/repo_language_stats.go b/modules/git/repo_language_stats.go
index 8ff8fa20c1..d623d6f57d 100644
--- a/modules/git/repo_language_stats.go
+++ b/modules/git/repo_language_stats.go
@@ -8,7 +8,6 @@ import (
 	"bytes"
 	"io"
 	"io/ioutil"
-	"math"
 
 	"code.gitea.io/gitea/modules/analyze"
 
@@ -21,7 +20,7 @@ import (
 const fileSizeLimit int64 = 16 * 1024 * 1024
 
 // GetLanguageStats calculates language stats for git repository at specified commit
-func (repo *Repository) GetLanguageStats(commitID string) (map[string]float32, error) {
+func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, error) {
 	r, err := git.PlainOpen(repo.Path)
 	if err != nil {
 		return nil, err
@@ -43,7 +42,6 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]float32, e
 	}
 
 	sizes := make(map[string]int64)
-	var total int64
 	err = tree.Files().ForEach(func(f *object.File) error {
 		if enry.IsVendor(f.Name) || enry.IsDotFile(f.Name) ||
 			enry.IsDocumentation(f.Name) || enry.IsConfiguration(f.Name) {
@@ -60,11 +58,10 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]float32, e
 
 		language := analyze.GetCodeLanguage(f.Name, content)
 		if language == enry.OtherLanguage || language == "" {
-			return nil
+			language = "other"
 		}
 
 		sizes[language] += f.Size
-		total += f.Size
 
 		return nil
 	})
@@ -72,21 +69,11 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]float32, e
 		return nil, err
 	}
 
-	stats := make(map[string]float32)
-	var otherPerc float32 = 100
-	for language, size := range sizes {
-		perc := float32(math.Round(float64(size)/float64(total)*1000) / 10)
-		if perc <= 0.1 {
-			continue
-		}
-		otherPerc -= perc
-		stats[language] = perc
+	if len(sizes) == 0 {
+		sizes["other"] = 0
 	}
-	otherPerc = float32(math.Round(float64(otherPerc)*10) / 10)
-	if otherPerc > 0 {
-		stats["other"] = otherPerc
-	}
-	return stats, nil
+
+	return sizes, nil
 }
 
 func readFile(f *object.File, limit int64) ([]byte, error) {
diff --git a/modules/indexer/stats/indexer_test.go b/modules/indexer/stats/indexer_test.go
index 29d0f6dbe4..b60c6d9bb4 100644
--- a/modules/indexer/stats/indexer_test.go
+++ b/modules/indexer/stats/indexer_test.go
@@ -34,6 +34,9 @@ func TestRepoStatsIndex(t *testing.T) {
 
 	repo, err := models.GetRepositoryByID(1)
 	assert.NoError(t, err)
+	status, err := repo.GetIndexerStatus(models.RepoIndexerTypeStats)
+	assert.NoError(t, err)
+	assert.Equal(t, "65f1bf27bc3bf70f64657658635e66094edbcb4d", status.CommitSha)
 	langs, err := repo.GetTopLanguageStats(5)
 	assert.NoError(t, err)
 	assert.Len(t, langs, 1)