You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Because I know that boost.parser uses utf32 internally, I specify all string literals as utf32 strings, like bp::lit(U"keyword").
Now I see in the profiler that quite some time is spent in utf_iterator<4,4,...>::read().
If FromFormat and ToFormat are identical, no transcoding is necessary. It would be nice if utf_iterator skipped all internal recoding if FromFormat and ToFormat are identical. Everything can be forwarded to curr().