diff --git a/README.md b/README.md
index 9c9b8162b35f4d2f7a289075f9c4fc8b9bbdf670..81ca6b6b2b77fa3f3e4dbebab4bbf7972ce87c09 100644
--- a/README.md
+++ b/README.md
@@ -1,3 +1,21 @@
+# Chinese OBFS
+
+This is a fork of BullshitGenerator.
+
+This tool could encode (hide) any data into nonsense chinese text, to avoid the extreme censorship of chinese forum. 
+
+This tool should be used to decode data from nonsense chinese text. 
+
+## Format proposal
+
+The first section encodes the topic. 
+
+The second section should use the bootstrap data.json and the topic above, to encode an url to actual data.json. 
+
+The third section is the actual data section. 
+
+----
+
 # 狗屁不通文章生成器
 # BullshitGenerator
 
diff --git a/bullshit.py b/bullshit.py
new file mode 100644
index 0000000000000000000000000000000000000000..346d886049c133252b9c56e7f7d012b7c08e9c71
--- /dev/null
+++ b/bullshit.py
@@ -0,0 +1,61 @@
+#!/usr/bin/python
+# -*- coding: UTF-8 -*-
+
+import os, re
+import random
+
+def loadJson(fileName):
+    import json
+    strList = fileName.split(".")
+    if strList[len(strList)-1].lower() == "json":
+        with open(fileName,mode='r',encoding="utf-8") as file:
+            return json.loads(file.read())
+
+data = loadJson("data.json")
+motto_data = data["famous"] # a 代表prefix_data，b代表postfix_data
+prefix_data = data["before"] # 在motto_data前面弄点nonsense_data
+postfix_data = data['after']  # 在motto_data后面弄点nonsense_data
+nonsense_data = data['bosh'] # 代表文章主要nonsense_data来源
+
+xx = "学生会退会"
+
+repeat_factor = 2
+
+def randomized_yield(iterable):
+    global repeat_factor
+    pool = list(iterable) * repeat_factor
+    while True:
+        random.shuffle(pool)
+        for ele in pool:
+            yield ele
+
+nonsense_generator = randomized_yield(nonsense_data)
+motto_generator = randomized_yield(motto_data)
+
+def new_motto():
+    global motto_generator
+    xx = next(motto_generator)
+    xx = xx.replace("a", random.choice(prefix_data) )
+    xx = xx.replace("b", random.choice(postfix_data) )
+    return xx
+
+def new_paragraph():
+    xx = ". "
+    xx += "\r\n"
+    xx += "    "
+    return xx
+
+if __name__ == "__main__":
+    xx = input("请输入文章主题:")
+    for x in xx:
+        tmp = str()
+        while ( len(tmp) < 6000 ) :
+            randsrc = random.randint(0,100)
+            if randsrc < 5:
+                tmp += new_paragraph()
+            elif randsrc < 20 :
+                tmp += new_motto()
+            else:
+                tmp += next(nonsense_generator)
+        tmp = tmp.replace("x",xx)
+        print(tmp)
diff --git a/readJSON.py b/readJSON.py
deleted file mode 100644
index 895b1fce06cc192b35cc847379f88e5308dd6ca4..0000000000000000000000000000000000000000
--- a/readJSON.py
+++ /dev/null
@@ -1,12 +0,0 @@
-# -*- coding: utf-8 -*-
-# @Author: Admin
-# @Date:   2019-11-01 16:52:34
-# @Last Modified by:   Admin
-# @Last Modified time: 2019-11-01 18:18:14
-def 读JSON文件(fileName=""):
-    import json
-    if fileName!='':
-        strList = fileName.split(".")
-        if strList[len(strList)-1].lower() == "json":
-            with open(fileName,mode='r',encoding="utf-8") as file:
-                return json.loads(file.read())
diff --git "a/\350\207\252\345\212\250\347\213\227\345\261\201\344\270\215\351\200\232\346\226\207\347\253\240\347\224\237\346\210\220\345\231\250.py" "b/\350\207\252\345\212\250\347\213\227\345\261\201\344\270\215\351\200\232\346\226\207\347\253\240\347\224\237\346\210\220\345\231\250.py"
deleted file mode 100644
index 2b63bfd949a9f360552fd426fbdd2e603328ecf8..0000000000000000000000000000000000000000
--- "a/\350\207\252\345\212\250\347\213\227\345\261\201\344\270\215\351\200\232\346\226\207\347\253\240\347\224\237\346\210\220\345\231\250.py"
+++ /dev/null
@@ -1,54 +0,0 @@
-#!/usr/bin/python
-# -*- coding: UTF-8 -*-
-
-import os, re
-import random,readJSON
-
-data = readJSON.读JSON文件("data.json")
-名人名言 = data["famous"] # a 代表前面垫话，b代表后面垫话
-前面垫话 = data["before"] # 在名人名言前面弄点废话
-后面垫话 = data['after']  # 在名人名言后面弄点废话
-废话 = data['bosh'] # 代表文章主要废话来源
-
-xx = "学生会退会"
-
-重复度 = 2
-
-def 洗牌遍历(列表):
-    global 重复度
-    池 = list(列表) * 重复度
-    while True:
-        random.shuffle(æ± )
-        for 元素 in 池:
-            yield 元素
-
-下一句废话 = 洗牌遍历(废话)
-下一句名人名言 = 洗牌遍历(名人名言)
-
-def 来点名人名言():
-    global 下一句名人名言
-    xx = next(下一句名人名言)
-    xx = xx.replace(  "a",random.choice(前面垫话) )
-    xx = xx.replace(  "b",random.choice(后面垫话) )
-    return xx
-
-def 另起一段():
-    xx = ". "
-    xx += "\r\n"
-    xx += "    "
-    return xx
-
-if __name__ == "__main__":
-    xx = input("请输入文章主题:")
-    for x in xx:
-        tmp = str()
-        while ( len(tmp) < 6000 ) :
-            分支 = random.randint(0,100)
-            if 分支 < 5:
-                tmp += 另起一段()
-            elif 分支 < 20 :
-                tmp += 来点名人名言()
-            else:
-                tmp += next(下一句废话)
-        tmp = tmp.replace("x",xx)
-        print(tmp)
\ No newline at end of file