From 9d652002c63d03d44083c4410881a457a9390e2f Mon Sep 17 00:00:00 2001
From: Cirno the Strongest <1447794+CirnoT@users.noreply.github.com>
Date: Sun, 31 May 2020 00:58:55 +0200
Subject: [PATCH] Fix language stat calculation (#11692)

* Fix language stat calculation

* Group languages and ignore 0 size files

* remove unneeded code
---
 models/repo_language_stats.go         | 35 +--------------------------
 modules/git/repo_language_stats.go    | 31 +++++++++++++++++++++---
 modules/indexer/stats/indexer_test.go |  4 +--
 3 files changed, 29 insertions(+), 41 deletions(-)

diff --git a/models/repo_language_stats.go b/models/repo_language_stats.go
index d08782eaf..a15063e25 100644
--- a/models/repo_language_stats.go
+++ b/models/repo_language_stats.go
@@ -26,22 +26,6 @@ type LanguageStat struct {
 	CreatedUnix timeutil.TimeStamp `xorm:"INDEX CREATED"`
 }
 
-// specialLanguages defines list of languages that are excluded from the calculation
-// unless they are the only language present in repository. Only languages which under
-// normal circumstances are not considered to be code should be listed here.
-var specialLanguages = map[string]struct{}{
-	"XML":      {},
-	"JSON":     {},
-	"TOML":     {},
-	"YAML":     {},
-	"INI":      {},
-	"SQL":      {},
-	"SVG":      {},
-	"Text":     {},
-	"Markdown": {},
-	"other":    {},
-}
-
 // LanguageStatList defines a list of language statistics
 type LanguageStatList []*LanguageStat
 
@@ -55,27 +39,12 @@ func (stats LanguageStatList) getLanguagePercentages() map[string]float32 {
 	langPerc := make(map[string]float32)
 	var otherPerc float32 = 100
 	var total int64
-	// Check that repository has at least one non-special language
-	var skipSpecial bool
-	for _, stat := range stats {
-		if _, ok := specialLanguages[stat.Language]; !ok {
-			skipSpecial = true
-			break
-		}
-	}
+
 	for _, stat := range stats {
-		// Exclude specific languages from percentage calculation
-		if _, ok := specialLanguages[stat.Language]; ok && skipSpecial {
-			continue
-		}
 		total += stat.Size
 	}
 	if total > 0 {
 		for _, stat := range stats {
-			// Exclude specific languages from percentage calculation
-			if _, ok := specialLanguages[stat.Language]; ok && skipSpecial {
-				continue
-			}
 			perc := float32(math.Round(float64(stat.Size)/float64(total)*1000) / 10)
 			if perc <= 0.1 {
 				continue
@@ -84,8 +53,6 @@ func (stats LanguageStatList) getLanguagePercentages() map[string]float32 {
 			langPerc[stat.Language] = perc
 		}
 		otherPerc = float32(math.Round(float64(otherPerc)*10) / 10)
-	} else {
-		otherPerc = 100
 	}
 	if otherPerc > 0 {
 		langPerc["other"] = otherPerc
diff --git a/modules/git/repo_language_stats.go b/modules/git/repo_language_stats.go
index d623d6f57..06d7d6aba 100644
--- a/modules/git/repo_language_stats.go
+++ b/modules/git/repo_language_stats.go
@@ -19,6 +19,20 @@ import (
 
 const fileSizeLimit int64 = 16 * 1024 * 1024
 
+// specialLanguages defines list of languages that are excluded from the calculation
+// unless they are the only language present in repository. Only languages which under
+// normal circumstances are not considered to be code should be listed here.
+var specialLanguages = []string{
+	"XML",
+	"JSON",
+	"TOML",
+	"YAML",
+	"INI",
+	"SVG",
+	"Text",
+	"Markdown",
+}
+
 // GetLanguageStats calculates language stats for git repository at specified commit
 func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, error) {
 	r, err := git.PlainOpen(repo.Path)
@@ -43,7 +57,7 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 
 	sizes := make(map[string]int64)
 	err = tree.Files().ForEach(func(f *object.File) error {
-		if enry.IsVendor(f.Name) || enry.IsDotFile(f.Name) ||
+		if f.Size == 0 || enry.IsVendor(f.Name) || enry.IsDotFile(f.Name) ||
 			enry.IsDocumentation(f.Name) || enry.IsConfiguration(f.Name) {
 			return nil
 		}
@@ -58,7 +72,13 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 
 		language := analyze.GetCodeLanguage(f.Name, content)
 		if language == enry.OtherLanguage || language == "" {
-			language = "other"
+			return nil
+		}
+
+		// group languages, such as Pug -> HTML; SCSS -> CSS
+		group := enry.GetLanguageGroup(language)
+		if group != "" {
+			language = group
 		}
 
 		sizes[language] += f.Size
@@ -69,8 +89,11 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 		return nil, err
 	}
 
-	if len(sizes) == 0 {
-		sizes["other"] = 0
+	// filter special languages unless they are the only language
+	if len(sizes) > 1 {
+		for _, language := range specialLanguages {
+			delete(sizes, language)
+		}
 	}
 
 	return sizes, nil
diff --git a/modules/indexer/stats/indexer_test.go b/modules/indexer/stats/indexer_test.go
index b60c6d9bb..4bcbaa942 100644
--- a/modules/indexer/stats/indexer_test.go
+++ b/modules/indexer/stats/indexer_test.go
@@ -39,7 +39,5 @@ func TestRepoStatsIndex(t *testing.T) {
 	assert.Equal(t, "65f1bf27bc3bf70f64657658635e66094edbcb4d", status.CommitSha)
 	langs, err := repo.GetTopLanguageStats(5)
 	assert.NoError(t, err)
-	assert.Len(t, langs, 1)
-	assert.Equal(t, "other", langs[0].Language)
-	assert.Equal(t, float32(100), langs[0].Percentage)
+	assert.Empty(t, langs)
 }