Limit codepoint caching to max 3-byte sequences

This keeps the benefit of avoiding sprintf on common codepoints without ballooning the lookup table too much.
avit · Nov 14, 2020 · 7781530 · 7781530
1 parent 0eb472e
commit 7781530
Showing 1 changed file with 3 additions and 1 deletion.
diff --git a/lib/rack/utf8_sanitizer.rb b/lib/rack/utf8_sanitizer.rb
@@ -250,9 +250,11 @@ def unescape_unreserved(input)
     # optimized from URI::RFC2396_Parser#escape
     def escape_unreserved(input)
       @unsafe_map ||= Hash.new do |table, us|
-        table[us] = us.each_byte.reduce('') do |tmp, uc|
+        encoded = us.each_byte.reduce('') do |tmp, uc|
           tmp << sprintf('%%%02X', uc)
         end
+        table[us] = encoded if us.bytesize <= 3
+        encoded
       end
       input.gsub(UNSAFE, @unsafe_map).force_encoding(Encoding::US_ASCII)
     end