The source project of this merge request has been removed.
Parse Unicode decimals as well as ASCII decimals when resolving page labels.
It isn't pretty but it does seem to work with the best and the worst. It does seem a bit convoluted to first convert this to a multibyte representation and then back to wide characters, but it is hard to follow these conversions when only a GooString
with or without a BOM is passed around. (Actually using std::string
and std::wstring
to handle UTF-8 and UTF-16 strings respectively would really be nice. But then again, Rust has me convinced that enforcing "everything must be UTF-8 always" via the type system is even nicer. I hear that even QString
is UTF-8 internally now...)