fix: use linguist-generated for language stats (#7685)

- Adds code comment to explain behavior of the `linguist-generated` gitattribute.
- Adjusts the code to ignore the file if `linguist-generated` is true.
- Resolves forgejo/forgejo#7677
- Adds unit testing.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/7685
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Co-authored-by: Gusted <postmaster@gusted.xyz>
Co-committed-by: Gusted <postmaster@gusted.xyz>

modules/git/repo_language_stats.go

@@ -204,7 +204,13 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
 				return nil, err
 			}
 		}
-		if !isTrue(isGenerated) && enry.IsGenerated(f.Name(), content) {
+
+		// We consider three cases:
+		// 1. linguist-generated=true, then we ignore the file.
+		// 2. linguist-generated=false, we don't ignore the file.
+		// 3. linguist-generated is not set, then `enry.IsGenerated` determines if the file is generated.
+		if isTrue(isGenerated) || !isFalse(isGenerated) && enry.IsGenerated(f.Name(), content) {
+			log.Trace("Ignore %q for language stats, because it is generated", f.Name())
 			continue
 		}
 

modules/git/repo_language_stats_test.go

@@ -25,6 +25,15 @@ func TestRepository_GetLanguageStats(t *testing.T) {
 		"Python": 134,
 		"Java":   112,
 	}, stats)
+
+	stats, err = gitRepo.GetLanguageStats("95d3505f2db273e40be79f84416051ae85e9ea0d")
+	require.NoError(t, err)
+
+	assert.Equal(t, map[string]int64{
+		"Cobra":  67,
+		"Python": 67,
+		"Java":   112,
+	}, stats)
 }
 
 func TestMergeLanguageStats(t *testing.T) {

modules/git/tests/repos/language_stats_repo/logs/HEAD

@@ -1,2 +0,0 @@
-0000000000000000000000000000000000000000 8fee858da5796dfb37704761701bb8e800ad9ef3 Andrew Thornton <art27@cantab.net> 1632140318 +0100	commit (initial): Add some test files for GetLanguageStats
-8fee858da5796dfb37704761701bb8e800ad9ef3 341fca5b5ea3de596dc483e54c2db28633cd2f97 oliverpool <git@olivier.pfad.fr> 1711278775 +0100	push

modules/git/tests/repos/language_stats_repo/logs/refs/heads/master

@@ -1,2 +0,0 @@
-0000000000000000000000000000000000000000 8fee858da5796dfb37704761701bb8e800ad9ef3 Andrew Thornton <art27@cantab.net> 1632140318 +0100	commit (initial): Add some test files for GetLanguageStats
-8fee858da5796dfb37704761701bb8e800ad9ef3 341fca5b5ea3de596dc483e54c2db28633cd2f97 oliverpool <git@olivier.pfad.fr> 1711278775 +0100	push

modules/git/tests/repos/language_stats_repo/packed-refs

@@ -0,0 +1,2 @@
+# pack-refs with: peeled fully-peeled sorted 
+95d3505f2db273e40be79f84416051ae85e9ea0d refs/heads/master

modules/git/tests/repos/language_stats_repo/refs/heads/master

@@ -1 +0,0 @@
-341fca5b5ea3de596dc483e54c2db28633cd2f97