gistfile1.rb

# Split unicode characters from supplementary planes (such as iOS 5 emoji) into surrogate pairs
# so that node.js can parse them.
def split_into_surrogate_pairs(text)
  text.gsub(/[\u{10000}-\u{10ffff}]/) do |match|
    codepoint = match.codepoints.first
    pair = [0xD7C0 + (codepoint >> 10), 0xDC00 | codepoint & 0x3FF]
    pair.pack("U*")
  end
end