wcwidth: provide our own implementation of wcwidth() and wcswidth()

This commit adds a new script, generate-wcwidth.py, that generates
wcwidth tables from the bundled files (Unicode 14.0):

* DerivedGeneralCategory.txt
* EastAsianWidth.txt

This commit also adds the functions my_wcwidth() and my_wcswidth()
that replaces the system’s wcwidth()+wcswidth(), and uses the
generated tables to map Unicode codepoints to widths.

This is inspired by both XTerm’s wcwidth implementation, as well as
https://github.com/jquast/wcwidth.

Both of those are based on/inspired by
https://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c

my-wcwidth.h

@@ -0,0 +1,14 @@
+#pragma once
+#include <wchar.h>
+
+#if FOOT_SYSTEM_WCWIDTH == 0
+
+int my_wcwidth(wchar_t wc);
+int my_wcswidth(const wchar_t *s, size_t n);
+
+#else
+
+static inline int my_wcwidth(wchar_t wc) { return wcwidth(wc); }
+static inline int my_wcswidth(const wchar_t *s, size_t n) { return wcswidth(s, n); }
+
+#endif