Support multi-byte character sets in config file

[tsjensen]

We already have Unicode and general multi-byte character set (MBCS) support for input text (from #1), but we'd also like to have that for the config file, so that we can design boxes based on Unicode box drawing characters.

At the moment, the config file is in ASCII format. It should be in UTF-8 format.

[tsjensen]

Just a small update on this topic. Things are slowly happening. The general roadmap tries to take the slow speed into consideration in order to avoid a feature branch that would need to live for years. It goes like this:

#	Activity	Status
1.	Add black-box tests	✔️
2.	Add unit tests Now we can perform major refactorings while being reasonably sure we don't break too many things.	✔️
3.	Add a bxstring module with a boxes abstraction for strings that can handle Unicode, invisible ANSI control characters, and characters that take up more than one column of text (bxstr_t). This makes use of libunistring.	✔️
4.	Enable the parser to parse UTF-8 files, and put the data read to use in our code base. When this is finished, the config file will be fully UTF-8 enabled, but you will still see only x instead of any non-ASCII character. Still, that's an important intermediate step.	✔️
5.	The internal boxes logic would still work with ASCII, but upon output, we would print the Unicode / ANSI enabled real box. We now have Unicode support for creating boxes.	✔️
6.	Add another test suite which applies the most common commands (create, mend, remove) to every design in the official config file ("sunny-day tests", as we use no special options).	✔️
7.	Add Unicode support for the other commands (remove box, mend box, ...)	✔️
8.	Get stuff to work on Windows. This is currently made difficult for lack of a proper terminal. Ideas?	✔️
9.	Extensive updates to the documentation will be required.	✔️
10.	Release a new version with all this.	🏃
11.	Lots more internal refactoring, so that we natively work with bxstr_t * instead of char *. Changing the string abstraction in a program whose only purpose is wrangling strings is a major change. Some of this is already done (remove, detect).	⌛

[✔️ = done, 🏃 = in progress, ⌛ = planned]

I'll update this comment whenever something fundamental changes. Last update: 2023-12-25

[tsjensen]

The feature branch things are happening on is color-unicode. There are already a few test cases with colored unicode box designs, and it's working! We've had some delays because our code base didn't easily allow major refactorings. Currently, I feel we need some more test cases to be confident enough to release something. Also, properly supporting Windows will be an issue, and a lot of required cleanup.

But, it's coming along!

[tsjensen]

[tsjensen]

Okay, we released v2.3.0 today with full UTF-8 support in config file! 🎉

Do give it a try!