refactor(linter): clean language names in primary detection

Update DetectPrimaryLanguage to normalize "C/C++" and "C++" to "c" for consistent counting.
Revise function comment and internal comments for clarity.
Remove redundant comment on finding most common language.

internal/linter/language.go

@@ -2,9 +2,7 @@ package linter
 
 import "strings"
 
-// DetectPrimaryLanguage returns the dominant language from the list of files.
+// DetectPrimaryLanguage returns a clean language name without slashes or special chars.
-// It counts detected languages and returns the most frequent one.
-// Falls back to "go" (common default) or "unknown".
 func DetectPrimaryLanguage(files []string) string {
 	if len(files) == 0 {
 		return "unknown"
@@ -15,6 +13,10 @@ func DetectPrimaryLanguage(files []string) string {
 		lang, err := DetectLanguage(file)
 		if err == nil && lang != nil {
 			name := strings.ToLower(lang.Name)
+			// Clean the name: turn "C/C++" into "c"
+			if name == "c/c++" || name == "c++" {
+				name = "c"
+			}
 			counts[name]++
 		}
 	}
@@ -23,7 +25,6 @@ func DetectPrimaryLanguage(files []string) string {
 		return "unknown"
 	}
 
-	// Find most common language
 	var bestLang string
 	maxCount := -1
 	for lang, count := range counts {
@@ -33,7 +34,7 @@ func DetectPrimaryLanguage(files []string) string {
 		}
 	}
 
-	// Bias toward Go in mixed repos
+	// Bias toward Go in mixed repos, otherwise prefer "c" for C files
 	if counts["go"] > 0 && counts["go"] >= maxCount/2 {
 		return "go"
 	}