-- -- lua-MessagePack : -- local r, jit = pcall(require, 'jit') if not r then jit = nil end local SIZEOF_NUMBER = string.pack and #string.pack('n', 0.0) or 8 local maxinteger local mininteger if not jit and _VERSION < 'Lua 5.3' then -- Lua 5.1 & 5.2 local loadstring = loadstring or load local luac = string.dump(loadstring "a = 1") local header = { luac:byte(1, 12) } SIZEOF_NUMBER = header[11] end local assert = assert local error = error local pairs = pairs local pcall = pcall local setmetatable = setmetatable local tostring = tostring local type = type local char = require'string'.char local format = require'string'.format local floor = require'math'.floor local tointeger = require'math'.tointeger or floor local frexp = require'math'.frexp or require'mathx'.frexp local ldexp = require'math'.ldexp or require'mathx'.ldexp local huge = require'math'.huge local tconcat = require'table'.concat local _ENV = nil local m = {} --[[ debug only local function hexadump (s) return (s:gsub('.', function (c) return format('%02X ', c:byte()) end)) end m.hexadump = hexadump --]] local function argerror (caller, narg, extramsg) error("bad argument #" .. tostring(narg) .. " to " .. caller .. " (" .. extramsg .. ")") end local function typeerror (caller, narg, arg, tname) argerror(caller, narg, tname .. " expected, got " .. type(arg)) end local function checktype (caller, narg, arg, tname) if type(arg) ~= tname then typeerror(caller, narg, arg, tname) end end local packers = setmetatable({}, { __index = function (t, k) if k == 1 then return end -- allows ipairs error("pack '" .. k .. "' is unimplemented") end }) m.packers = packers packers['nil'] = function (buffer) buffer[#buffer+1] = char(0xC0) -- nil end packers['boolean'] = function (buffer, bool) if bool then buffer[#buffer+1] = char(0xC3) -- true else buffer[#buffer+1] = char(0xC2) -- false end end packers['string_compat'] = function (buffer, str) local n = #str if n <= 0x1F then buffer[#buffer+1] = char(0xA0 + n) -- fixstr elseif n <= 0xFFFF then buffer[#buffer+1] = char(0xDA, -- str16 floor(n / 0x100), n % 0x100) elseif n <= 4294967295.0 then buffer[#buffer+1] = char(0xDB, -- str32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else error"overflow in pack 'string_compat'" end buffer[#buffer+1] = str end packers['_string'] = function (buffer, str) local n = #str if n <= 0x1F then buffer[#buffer+1] = char(0xA0 + n) -- fixstr elseif n <= 0xFF then buffer[#buffer+1] = char(0xD9, -- str8 n) elseif n <= 0xFFFF then buffer[#buffer+1] = char(0xDA, -- str16 floor(n / 0x100), n % 0x100) elseif n <= 4294967295.0 then buffer[#buffer+1] = char(0xDB, -- str32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else error"overflow in pack 'string'" end buffer[#buffer+1] = str end packers['binary'] = function (buffer, str) local n = #str if n <= 0xFF then buffer[#buffer+1] = char(0xC4, -- bin8 n) elseif n <= 0xFFFF then buffer[#buffer+1] = char(0xC5, -- bin16 floor(n / 0x100), n % 0x100) elseif n <= 4294967295.0 then buffer[#buffer+1] = char(0xC6, -- bin32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else error"overflow in pack 'binary'" end buffer[#buffer+1] = str end local set_string = function (str) if str == 'string_compat' then packers['string'] = packers['string_compat'] elseif str == 'string' then packers['string'] = packers['_string'] elseif str == 'binary' then packers['string'] = packers['binary'] else argerror('set_string', 1, "invalid option '" .. str .."'") end end m.set_string = set_string packers['map'] = function (buffer, tbl, n) if n <= 0x0F then buffer[#buffer+1] = char(0x80 + n) -- fixmap elseif n <= 0xFFFF then buffer[#buffer+1] = char(0xDE, -- map16 floor(n / 0x100), n % 0x100) elseif n <= 4294967295.0 then buffer[#buffer+1] = char(0xDF, -- map32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else error"overflow in pack 'map'" end for k, v in pairs(tbl) do packers[type(k)](buffer, k) packers[type(v)](buffer, v) end end packers['array'] = function (buffer, tbl, n) if n <= 0x0F then buffer[#buffer+1] = char(0x90 + n) -- fixarray elseif n <= 0xFFFF then buffer[#buffer+1] = char(0xDC, -- array16 floor(n / 0x100), n % 0x100) elseif n <= 4294967295.0 then buffer[#buffer+1] = char(0xDD, -- array32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else error"overflow in pack 'array'" end for i = 1, n do local v = tbl[i] packers[type(v)](buffer, v) end end local set_array = function (array) if array == 'without_hole' then packers['_table'] = function (buffer, tbl) local is_map, n, max = false, 0, 0 for k in pairs(tbl) do if type(k) == 'number' and k > 0 then if k > max then max = k end else is_map = true end n = n + 1 end if max ~= n then -- there are holes is_map = true end if is_map then packers['map'](buffer, tbl, n) else packers['array'](buffer, tbl, n) end end elseif array == 'with_hole' then packers['_table'] = function (buffer, tbl) local is_map, n, max = false, 0, 0 for k in pairs(tbl) do if type(k) == 'number' and k > 0 then if k > max then max = k end else is_map = true end n = n + 1 end if is_map then packers['map'](buffer, tbl, n) else packers['array'](buffer, tbl, max) end end elseif array == 'always_as_map' then packers['_table'] = function(buffer, tbl) local n = 0 for k in pairs(tbl) do n = n + 1 end packers['map'](buffer, tbl, n) end else argerror('set_array', 1, "invalid option '" .. array .."'") end end m.set_array = set_array packers['table'] = function (buffer, tbl) packers['_table'](buffer, tbl) end packers['unsigned'] = function (buffer, n) if n >= 0 then if n <= 0x7F then buffer[#buffer+1] = char(n) -- fixnum_pos elseif n <= 0xFF then buffer[#buffer+1] = char(0xCC, -- uint8 n) elseif n <= 0xFFFF then buffer[#buffer+1] = char(0xCD, -- uint16 floor(n / 0x100), n % 0x100) elseif n <= 4294967295.0 then buffer[#buffer+1] = char(0xCE, -- uint32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else buffer[#buffer+1] = char(0xCF, -- uint64 0, -- only 53 bits from double floor(n / 0x1000000000000) % 0x100, floor(n / 0x10000000000) % 0x100, floor(n / 0x100000000) % 0x100, floor(n / 0x1000000) % 0x100, floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) end else if n >= -0x20 then buffer[#buffer+1] = char(0x100 + n) -- fixnum_neg elseif n >= -0x80 then buffer[#buffer+1] = char(0xD0, -- int8 0x100 + n) elseif n >= -0x8000 then n = 0x10000 + n buffer[#buffer+1] = char(0xD1, -- int16 floor(n / 0x100), n % 0x100) elseif n >= -0x80000000 then n = 4294967296.0 + n buffer[#buffer+1] = char(0xD2, -- int32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else buffer[#buffer+1] = char(0xD3, -- int64 0xFF, -- only 53 bits from double floor(n / 0x1000000000000) % 0x100, floor(n / 0x10000000000) % 0x100, floor(n / 0x100000000) % 0x100, floor(n / 0x1000000) % 0x100, floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) end end end packers['signed'] = function (buffer, n) if n >= 0 then if n <= 0x7F then buffer[#buffer+1] = char(n) -- fixnum_pos elseif n <= 0x7FFF then buffer[#buffer+1] = char(0xD1, -- int16 floor(n / 0x100), n % 0x100) elseif n <= 0x7FFFFFFF then buffer[#buffer+1] = char(0xD2, -- int32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else buffer[#buffer+1] = char(0xD3, -- int64 0, -- only 53 bits from double floor(n / 0x1000000000000) % 0x100, floor(n / 0x10000000000) % 0x100, floor(n / 0x100000000) % 0x100, floor(n / 0x1000000) % 0x100, floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) end else if n >= -0x20 then buffer[#buffer+1] = char(0xE0 + 0x20 + n) -- fixnum_neg elseif n >= -0x80 then buffer[#buffer+1] = char(0xD0, -- int8 0x100 + n) elseif n >= -0x8000 then n = 0x10000 + n buffer[#buffer+1] = char(0xD1, -- int16 floor(n / 0x100), n % 0x100) elseif n >= -0x80000000 then n = 4294967296.0 + n buffer[#buffer+1] = char(0xD2, -- int32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) else buffer[#buffer+1] = char(0xD3, -- int64 0xFF, -- only 53 bits from double floor(n / 0x1000000000000) % 0x100, floor(n / 0x10000000000) % 0x100, floor(n / 0x100000000) % 0x100, floor(n / 0x1000000) % 0x100, floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100) end end end local set_integer = function (integer) if integer == 'unsigned' then packers['integer'] = packers['unsigned'] elseif integer == 'signed' then packers['integer'] = packers['signed'] else argerror('set_integer', 1, "invalid option '" .. integer .."'") end end m.set_integer = set_integer packers['float'] = function (buffer, n) local sign = 0 if n < 0.0 then sign = 0x80 n = -n end local mant, expo = frexp(n) if mant ~= mant then buffer[#buffer+1] = char(0xCA, -- nan 0xFF, 0x88, 0x00, 0x00) elseif mant == huge or expo > 0x80 then if sign == 0 then buffer[#buffer+1] = char(0xCA, -- inf 0x7F, 0x80, 0x00, 0x00) else buffer[#buffer+1] = char(0xCA, -- -inf 0xFF, 0x80, 0x00, 0x00) end elseif (mant == 0.0 and expo == 0) or expo < -0x7E then buffer[#buffer+1] = char(0xCA, -- zero sign, 0x00, 0x00, 0x00) else expo = expo + 0x7E mant = floor((mant * 2.0 - 1.0) * ldexp(0.5, 24)) buffer[#buffer+1] = char(0xCA, sign + floor(expo / 0x2), (expo % 0x2) * 0x80 + floor(mant / 0x10000), floor(mant / 0x100) % 0x100, mant % 0x100) end end packers['double'] = function (buffer, n) local sign = 0 if n < 0.0 then sign = 0x80 n = -n end local mant, expo = frexp(n) if mant ~= mant then buffer[#buffer+1] = char(0xCB, -- nan 0xFF, 0xF8, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00) elseif mant == huge or expo > 0x400 then if sign == 0 then buffer[#buffer+1] = char(0xCB, -- inf 0x7F, 0xF0, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00) else buffer[#buffer+1] = char(0xCB, -- -inf 0xFF, 0xF0, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00) end elseif (mant == 0.0 and expo == 0) or expo < -0x3FE then buffer[#buffer+1] = char(0xCB, -- zero sign, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00) else expo = expo + 0x3FE mant = floor((mant * 2.0 - 1.0) * ldexp(0.5, 53)) buffer[#buffer+1] = char(0xCB, sign + floor(expo / 0x10), (expo % 0x10) * 0x10 + floor(mant / 0x1000000000000), floor(mant / 0x10000000000) % 0x100, floor(mant / 0x100000000) % 0x100, floor(mant / 0x1000000) % 0x100, floor(mant / 0x10000) % 0x100, floor(mant / 0x100) % 0x100, mant % 0x100) end end local set_number = function (number) if number == 'float' then packers['number'] = function (buffer, n) if floor(n) == n and n < maxinteger and n > mininteger then packers['integer'](buffer, n) else packers['float'](buffer, n) end end elseif number == 'double' then packers['number'] = function (buffer, n) if floor(n) == n and n < maxinteger and n > mininteger then packers['integer'](buffer, n) else packers['double'](buffer, n) end end else argerror('set_number', 1, "invalid option '" .. number .."'") end end m.set_number = set_number for k = 0, 4 do local n = tointeger(2^k) local fixext = 0xD4 + k packers['fixext' .. tostring(n)] = function (buffer, tag, data) assert(#data == n, "bad length for fixext" .. tostring(n)) buffer[#buffer+1] = char(fixext, tag < 0 and tag + 0x100 or tag) buffer[#buffer+1] = data end end packers['ext'] = function (buffer, tag, data) local n = #data if n <= 0xFF then buffer[#buffer+1] = char(0xC7, -- ext8 n, tag < 0 and tag + 0x100 or tag) elseif n <= 0xFFFF then buffer[#buffer+1] = char(0xC8, -- ext16 floor(n / 0x100), n % 0x100, tag < 0 and tag + 0x100 or tag) elseif n <= 4294967295.0 then buffer[#buffer+1] = char(0xC9, -- ext&32 floor(n / 0x1000000), floor(n / 0x10000) % 0x100, floor(n / 0x100) % 0x100, n % 0x100, tag < 0 and tag + 0x100 or tag) else error"overflow in pack 'ext'" end buffer[#buffer+1] = data end function m.pack (data) local buffer = {} packers[type(data)](buffer, data) return tconcat(buffer) end local unpackers -- forward declaration local function unpack_cursor (c) local s, i, j = c.s, c.i, c.j if i > j then c:underflow(i) s, i, j = c.s, c.i, c.j end local val = s:byte(i) c.i = i+1 return unpackers[val](c, val) end m.unpack_cursor = unpack_cursor local function unpack_str (c, n) local s, i, j = c.s, c.i, c.j local e = i+n-1 if e > j or n < 0 then c:underflow(e) s, i, j = c.s, c.i, c.j e = i+n-1 end c.i = i+n return s:sub(i, e) end local function unpack_array (c, n) local t = {} for i = 1, n do t[i] = unpack_cursor(c) end return t end local function unpack_map (c, n) local t = {} for i = 1, n do local k = unpack_cursor(c) local val = unpack_cursor(c) if k == nil or k ~= k then k = m.sentinel end if k ~= nil then t[k] = val end end return t end local function unpack_float (c) local s, i, j = c.s, c.i, c.j if i+3 > j then c:underflow(i+3) s, i, j = c.s, c.i, c.j end local b1, b2, b3, b4 = s:byte(i, i+3) local sign = b1 > 0x7F local expo = (b1 % 0x80) * 0x2 + floor(b2 / 0x80) local mant = ((b2 % 0x80) * 0x100 + b3) * 0x100 + b4 if sign then sign = -1 else sign = 1 end local n if mant == 0 and expo == 0 then n = sign * 0.0 elseif expo == 0xFF then if mant == 0 then n = sign * huge else n = 0.0/0.0 end else n = sign * ldexp(1.0 + mant / 0x800000, expo - 0x7F) end c.i = i+4 return n end local function unpack_double (c) local s, i, j = c.s, c.i, c.j if i+7 > j then c:underflow(i+7) s, i, j = c.s, c.i, c.j end local b1, b2, b3, b4, b5, b6, b7, b8 = s:byte(i, i+7) local sign = b1 > 0x7F local expo = (b1 % 0x80) * 0x10 + floor(b2 / 0x10) local mant = ((((((b2 % 0x10) * 0x100 + b3) * 0x100 + b4) * 0x100 + b5) * 0x100 + b6) * 0x100 + b7) * 0x100 + b8 if sign then sign = -1 else sign = 1 end local n if mant == 0 and expo == 0 then n = sign * 0.0 elseif expo == 0x7FF then if mant == 0 then n = sign * huge else n = 0.0/0.0 end else n = sign * ldexp(1.0 + mant / 4503599627370496.0, expo - 0x3FF) end c.i = i+8 return n end local function unpack_uint8 (c) local s, i, j = c.s, c.i, c.j if i > j then c:underflow(i) s, i, j = c.s, c.i, c.j end local b1 = s:byte(i) c.i = i+1 return b1 end local function unpack_uint16 (c) local s, i, j = c.s, c.i, c.j if i+1 > j then c:underflow(i+1) s, i, j = c.s, c.i, c.j end local b1, b2 = s:byte(i, i+1) c.i = i+2 return b1 * 0x100 + b2 end local function unpack_uint32 (c) local s, i, j = c.s, c.i, c.j if i+3 > j then c:underflow(i+3) s, i, j = c.s, c.i, c.j end local b1, b2, b3, b4 = s:byte(i, i+3) c.i = i+4 return ((b1 * 0x100 + b2) * 0x100 + b3) * 0x100 + b4 end local function unpack_uint64 (c) local s, i, j = c.s, c.i, c.j if i+7 > j then c:underflow(i+7) s, i, j = c.s, c.i, c.j end local b1, b2, b3, b4, b5, b6, b7, b8 = s:byte(i, i+7) c.i = i+8 return ((((((b1 * 0x100 + b2) * 0x100 + b3) * 0x100 + b4) * 0x100 + b5) * 0x100 + b6) * 0x100 + b7) * 0x100 + b8 end local function unpack_int8 (c) local s, i, j = c.s, c.i, c.j if i > j then c:underflow(i) s, i, j = c.s, c.i, c.j end local b1 = s:byte(i) c.i = i+1 if b1 < 0x80 then return b1 else return b1 - 0x100 end end local function unpack_int16 (c) local s, i, j = c.s, c.i, c.j if i+1 > j then c:underflow(i+1) s, i, j = c.s, c.i, c.j end local b1, b2 = s:byte(i, i+1) c.i = i+2 if b1 < 0x80 then return b1 * 0x100 + b2 else return ((b1 - 0xFF) * 0x100 + (b2 - 0xFF)) - 1 end end local function unpack_int32 (c) local s, i, j = c.s, c.i, c.j if i+3 > j then c:underflow(i+3) s, i, j = c.s, c.i, c.j end local b1, b2, b3, b4 = s:byte(i, i+3) c.i = i+4 if b1 < 0x80 then return ((b1 * 0x100 + b2) * 0x100 + b3) * 0x100 + b4 else return ((((b1 - 0xFF) * 0x100 + (b2 - 0xFF)) * 0x100 + (b3 - 0xFF)) * 0x100 + (b4 - 0xFF)) - 1 end end local function unpack_int64 (c) local s, i, j = c.s, c.i, c.j if i+7 > j then c:underflow(i+7) s, i, j = c.s, c.i, c.j end local b1, b2, b3, b4, b5, b6, b7, b8 = s:byte(i, i+7) c.i = i+8 if b1 < 0x80 then return ((((((b1 * 0x100 + b2) * 0x100 + b3) * 0x100 + b4) * 0x100 + b5) * 0x100 + b6) * 0x100 + b7) * 0x100 + b8 else return ((((((((b1 - 0xFF) * 0x100 + (b2 - 0xFF)) * 0x100 + (b3 - 0xFF)) * 0x100 + (b4 - 0xFF)) * 0x100 + (b5 - 0xFF)) * 0x100 + (b6 - 0xFF)) * 0x100 + (b7 - 0xFF)) * 0x100 + (b8 - 0xFF)) - 1 end end function m.build_ext (tag, data) return nil end local function unpack_ext (c, n, tag) local s, i, j = c.s, c.i, c.j local e = i+n-1 if e > j or n < 0 then c:underflow(e) s, i, j = c.s, c.i, c.j e = i+n-1 end c.i = i+n return m.build_ext(tag, s:sub(i, e)) end unpackers = setmetatable({ [0xC0] = function () return nil end, [0xC2] = function () return false end, [0xC3] = function () return true end, [0xC4] = function (c) return unpack_str(c, unpack_uint8(c)) end, -- bin8 [0xC5] = function (c) return unpack_str(c, unpack_uint16(c)) end, -- bin16 [0xC6] = function (c) return unpack_str(c, unpack_uint32(c)) end, -- bin32 [0xC7] = function (c) return unpack_ext(c, unpack_uint8(c), unpack_int8(c)) end, [0xC8] = function (c) return unpack_ext(c, unpack_uint16(c), unpack_int8(c)) end, [0xC9] = function (c) return unpack_ext(c, unpack_uint32(c), unpack_int8(c)) end, [0xCA] = unpack_float, [0xCB] = unpack_double, [0xCC] = unpack_uint8, [0xCD] = unpack_uint16, [0xCE] = unpack_uint32, [0xCF] = unpack_uint64, [0xD0] = unpack_int8, [0xD1] = unpack_int16, [0xD2] = unpack_int32, [0xD3] = unpack_int64, [0xD4] = function (c) return unpack_ext(c, 1, unpack_int8(c)) end, [0xD5] = function (c) return unpack_ext(c, 2, unpack_int8(c)) end, [0xD6] = function (c) return unpack_ext(c, 4, unpack_int8(c)) end, [0xD7] = function (c) return unpack_ext(c, 8, unpack_int8(c)) end, [0xD8] = function (c) return unpack_ext(c, 16, unpack_int8(c)) end, [0xD9] = function (c) return unpack_str(c, unpack_uint8(c)) end, [0xDA] = function (c) return unpack_str(c, unpack_uint16(c)) end, [0xDB] = function (c) return unpack_str(c, unpack_uint32(c)) end, [0xDC] = function (c) return unpack_array(c, unpack_uint16(c)) end, [0xDD] = function (c) return unpack_array(c, unpack_uint32(c)) end, [0xDE] = function (c) return unpack_map(c, unpack_uint16(c)) end, [0xDF] = function (c) return unpack_map(c, unpack_uint32(c)) end, }, { __index = function (t, k) if k < 0xC0 then if k < 0x80 then return function (c, val) return val end elseif k < 0x90 then return function (c, val) return unpack_map(c, val % 0x10) end elseif k < 0xA0 then return function (c, val) return unpack_array(c, val % 0x10) end else return function (c, val) return unpack_str(c, val % 0x20) end end elseif k > 0xDF then return function (c, val) return val - 0x100 end else return function () error("unpack '" .. format('%#x', k) .. "' is unimplemented") end end end }) local function cursor_string (str) return { s = str, i = 1, j = #str, underflow = function () error "missing bytes" end, } end local function cursor_loader (ld) return { s = '', i = 1, j = 0, underflow = function (self, e) self.s = self.s:sub(self.i) e = e - self.i + 1 self.i = 1 self.j = 0 while e > self.j do local chunk = ld() if not chunk then error "missing bytes" end self.s = self.s .. chunk self.j = #self.s end end, } end function m.unpack (s) checktype('unpack', 1, s, 'string') local cursor = cursor_string(s) local data = unpack_cursor(cursor) if cursor.i <= cursor.j then error "extra bytes" end return data end function m.unpacker (src) if type(src) == 'string' then local cursor = cursor_string(src) return function () if cursor.i <= cursor.j then return cursor.i, unpack_cursor(cursor) end end elseif type(src) == 'function' then local cursor = cursor_loader(src) return function () if cursor.i > cursor.j then pcall(cursor.underflow, cursor, cursor.i) end if cursor.i <= cursor.j then return true, unpack_cursor(cursor) end end else argerror('unpacker', 1, "string or function expected, got " .. type(src)) end end set_string'string_compat' set_integer'unsigned' if SIZEOF_NUMBER == 4 then maxinteger = 16777215 mininteger = -maxinteger m.small_lua = true unpackers[0xCB] = nil -- double unpackers[0xCF] = nil -- uint64 unpackers[0xD3] = nil -- int64 set_number'float' else maxinteger = 9007199254740991 mininteger = -maxinteger set_number'double' if SIZEOF_NUMBER > 8 then m.long_double = true end end set_array'without_hole' m._VERSION = '0.5.2' m._DESCRIPTION = "lua-MessagePack : a pure Lua implementation" m._COPYRIGHT = "Copyright (c) 2012-2019 Francois Perrad" return m -- -- This library is licensed under the terms of the MIT/X11 license, -- like Lua itself. --