Add a documentation specifying the position of the libc project wrt wide chararacter and multibyte char support

[sivachandra]

We want to start with the position that:

1. Wide-char encoding will be that followed by the compiler. On most platforms it is UTF-32 and on Windows it is UTF-16.
2. Multi-byte to wide-char conversions assume utf-8 encoding of the multi-byte character strings irrespective of the current locale.
3. Wide-char to multi-byte conversions will convert to utf-8 encoded multi-byte strings irrespective of the current locale.
4. Explicit conversions will be as per the explicit expectation. For example, wide-char to char16_t string conversions will perform explicit conversion from the wide-char encoding to the UTF-16 encoding.

[llvmbot]

@llvm/issue-subscribers-libc