New features, functionality or any other improvements

[VladimirShaleev]

If you have any questions about missing functionality or requests for improvement, you can discuss them here.

Such questions may include:

• A similar Python library provides functionality that is missing here;
• Can improve the API or functionality of the library (for example, add the function ipaddress::make_ipv6_address("::%eth2") to simplify similar parsing ipaddress::ipv6_address::parse("::%eth2"), etc);
• Improving parsing algorithms or library architecture;
• And similar questions...

[VladimirShaleev]

I know what improvements would be nice to make. But within the limited time, the question that worries me, not least of all, is the Unicode encoding.

Currently, when parsing Unicode strings, truncation is used if possible (cast to char type). Or an error if the code point is greater than 127. Which, although incorrect, works in this specific case of parsing IP addresses.

But this introduces the slurred error problem. Because in case of an incorrect IP address, instead of seeing the code of the incorrect Unicode character, we will see random characters that appeared after truncation.

I propose that when analyzing Unicode strings, analyze incoming encodings and, in the case of invalid characters, generate an error with the value of this Unicode code.

Details in this issue

[VladimirShaleev]

Added Unicode support for wide chars (wchar_t), utf-8, utf-16 and utf-32