module ActiveSupport::Multibyte::Unicode

Experimental RBS support (using type sampling data from the type_fusion project). 

# sig/active_support/multibyte/unicode.rbs

module ActiveSupport::Multibyte::Unicode
  def tidy_bytes: (String string, ?false force) -> String
end

def compose(codepoints)

Compose decomposed characters to the composed form. 
def compose(codepoints)
  codepoints.pack("U*").unicode_normalize(:nfc).codepoints
end

def decompose(type, codepoints)

Decompose composed characters to the decomposed form. 
def decompose(type, codepoints)
  if type == :compatibility
    codepoints.pack("U*").unicode_normalize(:nfkd).codepoints
  else
    codepoints.pack("U*").unicode_normalize(:nfd).codepoints
  end
end

def recode_windows1252_chars(string) 

def recode_windows1252_chars(string)
  string.encode(Encoding::UTF_8, Encoding::Windows_1252, invalid: :replace, undef: :replace)
end

def tidy_bytes(string, force = false)

Experimental RBS support (using type sampling data from the type_fusion project). 

def tidy_bytes: (String string, ?false force) -> String

This signature was generated using 1 sample from 1 application.

encoding is entirely CP1252 or ISO-8859-1.
Passing +true+ will forcibly tidy all bytes, assuming that the string's

resulting in a valid UTF-8 string.
Replaces all ISO-8859-1 or CP1252 characters by their UTF-8 equivalent 
def tidy_bytes(string, force = false)
  return string if string.empty? || string.ascii_only?
  return recode_windows1252_chars(string) if force
  string.scrub { |bad| recode_windows1252_chars(bad) }
end