Question

I use the Boost library to implement my application. All the string characters in the data model of my application are wide chars (wchar_t type). But in the boost library, some classes only hand the narrow char (char type), i.e. “address boost::asio::ip::address::from_string(const char* str)“. So I need to make the conversion between std::string and std::wstring when call the boost functions.

1. Is there performance issue due to the string conversions?
2. In there the converter in Boost, which makes the conversion between std::wstring and std::string with good performance?

UPDATE

Regarding the converter function. I find the code below works.

std::wstring wstr(L"Hello World");
const std::string nstr( wstr.begin(),  wstr.end());
const std::wstring wstr2(nstr.begin(), nstr.end());

Answer 1

Add the research conclusion myself.

Regarding the performance overhead of the string conversion. I debugged into the functions above. The conversion is implemented by the C-cast char by char. The time complexity is O(L), L is the length of the string. In my application, the strings required to be converted are not very long. So I don’t think there is any obviously performance latency due to the conversions.