Soften memory pressure

lib/linguist/generated.rb

@@ -241,22 +241,26 @@ module Linguist
       return lines[0].include?("Code generated by")
     end
 
+    PROTOBUF_EXTENSIONS = ['.py', '.java', '.h', '.cc', '.cpp']
+
     # Internal: Is the blob a C++, Java or Python source file generated by the
     # Protocol Buffer compiler?
     #
     # Returns true of false.
     def generated_protocol_buffer?
-      return false unless ['.py', '.java', '.h', '.cc', '.cpp'].include?(extname)
+      return false unless PROTOBUF_EXTENSIONS.include?(extname)
       return false unless lines.count > 1
 
       return lines[0].include?("Generated by the protocol buffer compiler.  DO NOT EDIT!")
     end
 
+    APACHE_THRIFT_EXTENSIONS = ['.rb', '.py', '.go', '.js', '.m', '.java', '.h', '.cc', '.cpp']
+
     # Internal: Is the blob generated by Apache Thrift compiler?
     #
     # Returns true or false
     def generated_apache_thrift?
-      return false unless ['.rb', '.py', '.go', '.js', '.m', '.java', '.h', '.cc', '.cpp'].include?(extname)
+      return false unless APACHE_THRIFT_EXTENSIONS.include?(extname)
       return false unless lines.count > 1
 
       return lines[0].include?("Autogenerated by Thrift Compiler") || lines[1].include?("Autogenerated by Thrift Compiler")

lib/linguist/heuristics.rb

@@ -56,7 +56,8 @@ module Linguist
 
     # Internal: Check if this heuristic matches the candidate languages.
     def matches?(filename)
-      @extensions.any? { |ext| filename.downcase.end_with?(ext) }
+      filename = filename.downcase
+      @extensions.any? { |ext| filename.end_with?(ext) }
     end
 
     # Internal: Perform the heuristic

lib/linguist/lazy_blob.rb

@@ -79,6 +79,10 @@ module Linguist
       @size
     end
 
+    def cleanup!
+      @data.clear if @data
+    end
+
     protected
 
     # Returns true if the attribute is present and not the string "false".

lib/linguist/repository.rb

@@ -157,8 +157,11 @@ module Linguist
 
           blob = Linguist::LazyBlob.new(repository, delta.new_file[:oid], new, mode.to_s(8))
 
-          next unless blob.include_in_language_stats?
-          file_map[new] = [blob.language.group.name, blob.size]
+          if blob.include_in_language_stats?
+            file_map[new] = [blob.language.group.name, blob.size]
+          end
+
+          blob.cleanup!
         end
       end