4 ms·
UTF-32 performance woes are exaggerated. UTF-8 decoding requires branching whereas UTF-32 does not and in terms of storage real world text does not consume much
by hgs3 4y ago
UTF-32 performance woes are exaggerated. UTF-8 decoding requires branching whereas UTF-32 does not and in terms of storage real world text does not consume much space.
- kzrdude 4y agoUtf32 has the BOM which adds complexity and at least a top level branch. Maybe it's not so bad in practice, just duplicated generated code. But as an internal representation, some of it can be ignored by choice.