mirror of
https://github.com/ruby/ruby.git
synced 2026-10-29 17:06:21 +08:00
The buffer is copied over based on the new string's termlen so we can just set the new encoding raw instead of recalculating based on the old encoding's termlen.