This is a benchmark of the proposed new utf8 encoding strategy for node.js. The main reason for this is that node needs to have its "own" utf-8 encoder that properly handles surrogate pairs.
new
- Uses the proposed new utf8 encoder andv8::String::Memory
api.old_nohint
- Usesv8::String::WriteUTF8
without the HINT_MANY_WRITES_EXPECTED flag.old_hint
- Usesv8::String::WriteUTF8
with the HINT_MANY_WRITES_EXPECTED