Joseph cheng indiejoseph

## pos.text
语料库在线 http://www.cncorpus.org

n   名词
nt  时间名词
nd  方位名词
nl  处所名词
nh  人名
nhf 姓
nhg 名
nn  族名

## gist:703422a3583d841fb482
http://zh.wikipedia.org/w/api.php?action=query&titles=%E8%B4%9D%E5%A1%9E%E5%B0%94%E6%9B%B2%E7%BA%BF&redirects=&converttitles=&prop=revisions&rvprop=content&format=json

## dbc2sbc.coffee
DBC2SBC = (str, flag) ->
  result=''
  return no  if str.length <= 0

  for i in [0...str.length]
    str1=str.charCodeAt(i)
    if !flag
      if str1 < 127
        result += String.fromCharCode str.charCodeAt(i) + 65248
      else

## ictclas_pos.py
POS = {
  "n": {  #1. 名词  (1个一类，7个二类，5个三类)
    "n":"名词",
    "nr":"人名",
    "nr1":"汉语姓氏",
    "nr2":"汉语名字",
    "nrj":"日语人名",
    "nrf":"音译人名",
    "ns":"地名",
    "nsf":"音译地名",

## s2t.txt
台 臺
啓 啟
老板 老闆
開髮 開發
爲 為
裏 裡
衆 眾

## BaumWelch.coffee
'use strict'

_ = require 'lodash'

MIN_FLOAT = -3.14e100

Object::default = (prop, value) ->
  @[prop] = value  unless @hasOwnProperty(prop)

Object::getValue = (prop, value) ->

## log_add.coffee
logAdd = (x, y) ->
  maximum = Math.max x, y
  minimum = Math.min x, y
  return maximum  if Math.abs(maximum - minimum) > 30
  return maximum + Math.log 1 + Math.exp(maximum - minimum)

## sort_object.coffee
obj = do (obj) ->
  res = {}
  keys = Object.keys(obj).sort (a, b) -> obj[b] - obj[a]
  res[name] = obj[name] for name in keys
  res

## helpers.coffee
sum = (arr) -> arr.reduce ((a,b) -> a+b), 0
cmp = (x, y) -> (if x > y then 1 else (if x < y then -1 else 0))
keys = Object.keys(list).sort((a,b)-> list[b]-list[a]) # sort by object value DESC

# initial array with default value
_.range(3).map(function () { return 'a' })

# string to bigram array
toBigrams = (str) ->
  oneGrams = str.split('')

## 现代汉语常用词表.txt
阿爸	a1'ba4	18137
阿昌族	a1'chang1'zu2	50849
阿斗	a1'dou3	42632
阿飞	a1'fei1	48603
阿富汗	a1'fu4'han4	3461
阿訇	a1'hong1	34432
阿拉伯数字	a1'la1'bo2'shu4'zi4	35937
阿拉伯语	a1'la1'bo2'yu3	30476
阿妈	a1'ma1	16220
	语料库在线 http://www.cncorpus.org

	n 名词
	nt 时间名词
	nd 方位名词
	nl 处所名词
	nh 人名
	nhf 姓
	nhg 名
	nn 族名
	DBC2SBC = (str, flag) ->
	result=''
	return no if str.length <= 0

	for i in [0...str.length]
	str1=str.charCodeAt(i)
	if !flag
	if str1 < 127
	result += String.fromCharCode str.charCodeAt(i) + 65248
	else
	POS = {
	"n": { #1. 名词 (1个一类，7个二类，5个三类)
	"n":"名词",
	"nr":"人名",
	"nr1":"汉语姓氏",
	"nr2":"汉语名字",
	"nrj":"日语人名",
	"nrf":"音译人名",
	"ns":"地名",
	"nsf":"音译地名",
	'use strict'

	_ = require 'lodash'

	MIN_FLOAT = -3.14e100

	Object::default = (prop, value) ->
	@[prop] = value unless @hasOwnProperty(prop)

	Object::getValue = (prop, value) ->
	logAdd = (x, y) ->
	maximum = Math.max x, y
	minimum = Math.min x, y
	return maximum if Math.abs(maximum - minimum) > 30
	return maximum + Math.log 1 + Math.exp(maximum - minimum)
	obj = do (obj) ->
	res = {}
	keys = Object.keys(obj).sort (a, b) -> obj[b] - obj[a]
	res[name] = obj[name] for name in keys
	res
	sum = (arr) -> arr.reduce ((a,b) -> a+b), 0
	cmp = (x, y) -> (if x > y then 1 else (if x < y then -1 else 0))
	keys = Object.keys(list).sort((a,b)-> list[b]-list[a]) # sort by object value DESC

	# initial array with default value
	_.range(3).map(function () { return 'a' })

	# string to bigram array
	toBigrams = (str) ->
	oneGrams = str.split('')
	阿爸 a1'ba4 18137
	阿昌族 a1'chang1'zu2 50849
	阿斗 a1'dou3 42632
	阿飞 a1'fei1 48603
	阿富汗 a1'fu4'han4 3461
	阿訇 a1'hong1 34432
	阿拉伯数字 a1'la1'bo2'shu4'zi4 35937
	阿拉伯语 a1'la1'bo2'yu3 30476
	阿妈 a1'ma1 16220