Awesome

AbstractCodec

The abstract codec class to encode/decode buffer/string.

• Codec
  • name: the codec name.
  • options: the options passed via the encode/decode.
  • encode(value, options): encode the value.
    • return the encoded string, or encoded buffer if options.buffer is true
      • note: the return encoded buffer is a global buffer instance on the codec.
    • return the byte length of encoded value if options.buffer is true or is a Buffer.
    • options.encoding (string or codec instance): return the value directly if no encoding
    • options.buffer: the destBuffer or true.
      • options.bufferEncoding (string): the Buffer encoding used via Buffer. defaults to 'utf8'
      • options.bufferOffset (int): the offset of destBuffer. defaults to 0. if options.buffer is a Buffer.
  • decode(value, options): decode the value.
    • return the decoded value.
    • options.encoding (string or codec instance): return the value directly if no encoding
    • if value is Buffer:
      • options.bufferEncoding (string): the Buffer encoding used via value is Buffer. defaults to 'utf8'
      • options.bufferStart (int): the start of value. defaults to 0.
      • options.bufferEnd (int): the end of value. defaults to value.length - options.bufferStart.
  • encodeString(value): encode the value. return the encoded string.
  • decodeString(aString): decode the string(value). return the decoded value.
  • encodeBuffer(value, destBuffer, offset=0, encoding='utf8'):
    • encode value to the destBuffer. return the encoded length.
    • it just return the encoded byte length if the destBuffer is null
    • the default start is 0 offset of destBuffer.
  • decodeBuffer(buffer, start, end, encoding='utf8'):
    • decode the buffer. return the decoded value.
    • the default start is 0, end is buffer.length - start.
  • buffer: the Buffer instance.
    • it's avaiable only when constructor passed bufferSize argument or _encodeBuffer implenmented only.
  • bufferSize: the default max interal buffer size.
  • isBuffer(): it's true if have a interal buffer.
  • byteLength(value): return the byte length of the value.

Codec Usage

var Codec = require("buffer-codec")
var json = Codec("json")

var data = {a:1,b:2}
var encodedData = json.encode(data)

assert.equal(json.decode(encodedData), data)

Develope A New Codec

you should implenment:

• _encodeString/_decodeString or _encodeBuffer/_decodeBuffer
  • _encodeString(value): encode the value. return the encoded string.
  • _decodeString(value): decode the string(value). return the decoded value.
  • _encodeBuffer(value, destBuffer, offset=0, encoding='utf8'):
    • encode value to the destBuffer. return the encoded length.
    • it just return the encoded byte length if the destBuffer is null
    • the default start is 0 offset of destBuffer.
  • _decodeBufer(buffer, start, end, encoding='utf8'):
    • decode the buffer. return the decoded value.
    • the default start is 0, end is buffer.length - start.

Codec = require("buffer-codec")
register = Codec.register
aliases  = Codec.aliases

class TextCodec
  register TextCodec
  aliases  TextCodec, 'utf8', 'utf-8'

  _encodeString: (data)->
    if not data? or Buffer.isBuffer data
      data
    else
      String(data)
  _decodeString: (data)->data
  byteLength: (data)->
    if data? then data.length || String(data).length else 0

class JsonCodec
  register JsonCodec, TextCodec

  constructor: -> return super
  _encodeString: JSON.stringify
  _decodeString: JSON.parse
  byteLength: (data)->@_encodeString(data).length


# Using:

# get the JsonCodec Class
# lowercase name only here:
JsonCodec = Codec['json']
# or
JsonCodec = TextCodec['json']

# get the global JsonCodec instance from the Codec
json = Codec('json')
# or:
json = JsonCodec()

JsonCodec().should.be.equal Codec('json')

# create a new JsonCodec instance.
json2 = new JsonCodec()

json2.should.not.be.equal json

# reuse this buffer instead of create every once.
buf = new Buffer(8192)

bufLen = json.encodeBuffer({a:1,b:2}, buf)

data = json.decodeBuffer(buf, 0, bufLen)

Codec List:

• Text Codec: encode via toString() , decode return the data directly.
  • Json Codec: encode via JSON.stringify(.toJSON), decode via JSON.parse
  • Hex Codec: hex string to Binary.
• Binary Codec:
  • encodeBuffer: encode string or array to a buffer.
  • decodeBuffer: return the buffer directly.
  • encodeString:
    • result is the same string if value is string
    • result is ascii string if value is array, the number element in array saved is (element & 0xFF) if element is not number, saved 0 instead.
  • decodeString: return the same string.
• Bytewise Codec package: A binary string serialization which sorts bytewise for arbitrarily complex data structures, respecting bytewise structured sorting efficiently.
  • buffer-codec-bytewise
  • npm install buffer-codec-bytewise