Add test cases for Utils.utf8_encode

DataDog · Jan 16, 2018 · d8385df · d8385df
1 parent f912cba
commit d8385df
Show file tree

Hide file tree

Showing 2 changed files with 32 additions and 1 deletion.
diff --git a/lib/ddtrace/utils.rb b/lib/ddtrace/utils.rb
@@ -51,6 +51,6 @@ def self.utf8_encode(str, options = {})
       options[:placeholder] || STRING_PLACEHOLDER
     end
 
-    STRING_PLACEHOLDER = ''.freeze
+    STRING_PLACEHOLDER = ''.encode(Encoding::UTF_8).freeze
   end
 end
diff --git a/test/utils_test.rb b/test/utils_test.rb
@@ -58,4 +58,35 @@ def test_forked_process_id_collision
     refute_equal(Datadog::Utils.next_id, r.read.chomp.to_i)
     r.close
   end
+
+  def test_utf8_encoding_happy_path
+    str = 'pristine ✓'
+
+    assert_equal('pristine ✓', Datadog::Utils.utf8_encode(str))
+
+    assert_equal(::Encoding::UTF_8, Datadog::Utils.utf8_encode(str).encoding)
+
+    # we don't allocate new objects when a valid UTF-8 string is provided
+    assert_same(str, Datadog::Utils.utf8_encode(str))
+  end
+
+  def test_utf8_encoding_invalid_conversion
+    time_bomb = "\xC2".force_encoding(::Encoding::ASCII_8BIT)
+
+    # making sure this is indeed a problem
+    assert_raises(Encoding::CompatibilityError) do
+      time_bomb.encode(Encoding::UTF_8)
+    end
+
+    assert_equal(Datadog::Utils::STRING_PLACEHOLDER, Datadog::Utils.utf8_encode(time_bomb))
+
+    # we can also set a custom placeholder
+    assert_equal('?', Datadog::Utils.utf8_encode(time_bomb, placeholder: '?'))
+  end
+
+  def test_binary_data
+    byte_array = "keep what \xC2 is valid".force_encoding(::Encoding::ASCII_8BIT)
+
+    assert_equal('keep what is valid', Utils.utf8_encode(byte_array, binary: true))
+  end
 end