enc 0.2.2

• Remove clever code that breaks enc on R-devel.

• Use LEVELS macro instead of low-level access of opaque structure (#23).
• Fix write_lines_enc() on Windows systems: input is always transformed to UTF-8 before writing (#21).
• Fix test on systems with latin-1 encoding.
• Remove documentation link absent in current rlang.

• Fix transformation of non-ASCII text on Windows (#18, @yutannihilation).

Initial release.

• utf8 class with constructor, coercion, combination, formatting, printing, and checked updates.
• to_encoding() performs deep encoding conversion of objects, including names and other attributes. Variants: to_utf8(), to_native(), to_latin1() and to_alien().
• encoding(), returns "ASCII" for pure ASCII strings and behaves identically to base::Encoding() otherwise.
• all_utf8(), returns a logical scalar that indicates if all elements of a character vector are UTF-8 encoded; this includes pure ASCII strings.
• read_lines_enc(), try_read_lines_enc(), and write_lines_enc() for robust reading and writing of text files. Returns/accepts objects of class utf8.
• transform_lines_enc(), with robust handling if only some files could be transformed in transform_lines_enc(). Uses try_read_lines_enc(), therefore only warns if file is missing. Auto-detects and maintains EOL delimiter.