Unicode

Evo strings and I/O streams are byte-based and are naturally compatible with ASCII and UTF-8.

• Linux/Unix APIs prefer UTF-8 for Unicode
• Windows APIs prefer UTF-16 for Unicode
  • Note that the Windows console doesn't normally display UTF-8 correctly
• Evo defines it's own consistent wide character types (all unsigned):
  • wchar16 – UTF-16 character
  • wchar32 – UTF-32 character, full code point
• Evo doesn't use the C/C++ wchar_t type since it's platform dependent
• C++11: Evo has support for UTF-16 string literals (u prefix, via char16_t), see UnicodeString and String

Recommended: Use String (UTF-8) and convert with UnicodeString (UTF-16) on demand as needed for APIs that require UTF-16 (like Windows).

For background and rationale see: http://utf8everywhere.org

Conversion

• Convert to UTF-8 with String::set_unicode()
• Convert to UTF-16 with UnicodeString

Full Unicode with ICU

ICU4C (i.e. libicu) is a mature C/C++ library for Unicode support and software globalization.

• Evo integration with libicu coming soon...
• For more on libicu: http://site.icu-project.org

Low-Level Functions

These are implemented with Evo and don't require libicu.

• utf8_scan()
• utf8_min()
• utf8_count()
• utf8_to16()
• utf16_scan()
• utf16_scan_term()
• utf16_compare()
• utf16_strlen()
• utf16_min()
• utf16_count()
• utf16_to8()