Decode a string
encoded in Punycode
# File lib/domain_name/punycode.rb, line 198 def decode(string) # Initialize the state n = INITIAL_N i = 0 bias = INITIAL_BIAS if j = string.rindex(DELIMITER) b = string[0...j] b.match(RE_NONBASIC) and raise ArgumentError, "Illegal character is found in basic part: #{string.inspect}" # Handle the basic code points output = b.unpack('U*') u = string[(j + 1)..-1] else output = [] u = string end # Main decoding loop: Start just after the last delimiter if any # basic code points were copied; start at the beginning # otherwise. input = u.unpack('C*') input_length = input.length h = 0 out = output.length while h < input_length # Decode a generalized variable-length integer into delta, # which gets added to i. The overflow checking is easier # if we increase i as we go, then subtract off its starting # value at the end to obtain delta. oldi = i w = 1 k = BASE loop { digit = decode_digit(input[h]) or raise ArgumentError, "Illegal character is found in non-basic part: #{string.inspect}" h += 1 i += digit * w raise BufferOverflowError if i > MAXINT t = k <= bias ? TMIN : k - bias >= TMAX ? TMAX : k - bias break if digit < t w *= BASE - t raise BufferOverflowError if w > MAXINT k += BASE h < input_length or raise ArgumentError, "Malformed input given: #{string.inspect}" } # Adapt the bias delta = oldi == 0 ? i / DAMP : (i - oldi) >> 1 delta += delta / (out + 1) bias = 0 while delta > CUTOFF delta /= LOBASE bias += BASE end bias += (LOBASE + 1) * delta / (delta + SKEW) # i was supposed to wrap around from out+1 to 0, incrementing # n each time, so we'll fix that now: q, i = i.divmod(out + 1) n += q raise BufferOverflowError if n > MAXINT # Insert n at position i of the output: output[i, 0] = n out += 1 i += 1 end output.pack('U*') end
Decode a hostname using IDN/Punycode algorithms
# File lib/domain_name/punycode.rb, line 280 def decode_hostname(hostname) hostname.gsub(%r(\A|\.)xn--([^.]*)/) { $1 << decode($2) } end
Encode a string
in Punycode
# File lib/domain_name/punycode.rb, line 105 def encode(string) input = string.unpack('U*') output = '' # Initialize the state n = INITIAL_N delta = 0 bias = INITIAL_BIAS # Handle the basic code points input.each { |cp| output << cp.chr if cp < 0x80 } h = b = output.length # h is the number of code points that have been handled, b is the # number of basic code points, and out is the number of characters # that have been output. output << DELIMITER if b > 0 # Main encoding loop while h < input.length # All non-basic code points < n have been handled already. Find # the next larger one m = MAXINT input.each { |cp| m = cp if (n...m) === cp } # Increase delta enough to advance the decoder's <n,i> state to # <m,0>, but guard against overflow delta += (m - n) * (h + 1) raise BufferOverflowError if delta > MAXINT n = m input.each { |cp| # AMC-ACE-Z can use this simplified version instead if cp < n delta += 1 raise BufferOverflowError if delta > MAXINT elsif cp == n # Represent delta as a generalized variable-length integer q = delta k = BASE loop { t = k <= bias ? TMIN : k - bias >= TMAX ? TMAX : k - bias break if q < t q, r = (q - t).divmod(BASE - t) output << encode_digit(t + r, false) k += BASE } output << encode_digit(q, false) # Adapt the bias delta = h == b ? delta / DAMP : delta >> 1 delta += delta / (h + 1) bias = 0 while delta > CUTOFF delta /= LOBASE bias += BASE end bias += (LOBASE + 1) * delta / (delta + SKEW) delta = 0 h += 1 end } delta += 1 n += 1 end output end
Encode a hostname using IDN/Punycode algorithms
# File lib/domain_name/punycode.rb, line 185 def encode_hostname(hostname) hostname.match(RE_NONBASIC) or return hostname hostname.split('.').map { |name| if name.match(RE_NONBASIC) 'xn--' << encode(name) else name end }.join('.') end