{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# json 模块：处理 JSON 数据"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[JSON (JavaScript Object Notation)](http://json.org) 是一种轻量级的数据交换格式，易于人阅读和编写，同时也易于机器解析和生成。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## JSON 基础"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`JSON` 的基础结构有两种：键值对 (`name/value pairs`) 和数组 (`array`)。\n",
    "\n",
    "`JSON` 具有以下形式：\n",
    "\n",
    "- `object` - 对象，用花括号表示，形式为（数据是无序的）：\n",
    "    - `{ pair_1, pair_2, ..., pair_n }`\n",
    "- `pair` - 键值对，形式为：\n",
    "    - `string : value`\n",
    "- `array` - 数组，用中括号表示，形式为（数据是有序的）：\n",
    "    - `[value_1, value_2, ..., value_n ]`\n",
    "- `value` - 值，可以是\n",
    "    - `string` 字符串\n",
    "    - `number` 数字\n",
    "    - `object` 对象\n",
    "    - `array` 数组\n",
    "    - `true / false / null` 特殊值\n",
    "- `string` 字符串\n",
    "\n",
    "例子：\n",
    "\n",
    "```json\n",
    "{\n",
    "    \"name\": \"echo\",\n",
    "    \"age\": 24,\n",
    "    \"coding skills\": [\"python\", \"matlab\", \"java\", \"c\", \"c++\", \"ruby\", \"scala\"],\n",
    "    \"ages for school\": { \n",
    "        \"primary school\": 6,\n",
    "        \"middle school\": 9,\n",
    "        \"high school\": 15,\n",
    "        \"university\": 18\n",
    "    },\n",
    "    \"hobby\": [\"sports\", \"reading\"],\n",
    "    \"married\": false\n",
    "}\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## JSON 与 Python 的转换"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "假设我们已经将上面这个 `JSON` 对象写入了一个字符串："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import json\n",
    "from pprint import pprint\n",
    "\n",
    "info_string = \"\"\"\n",
    "{\n",
    "    \"name\": \"echo\",\n",
    "    \"age\": 24,\n",
    "    \"coding skills\": [\"python\", \"matlab\", \"java\", \"c\", \"c++\", \"ruby\", \"scala\"],\n",
    "    \"ages for school\": { \n",
    "        \"primary school\": 6,\n",
    "        \"middle school\": 9,\n",
    "        \"high school\": 15,\n",
    "        \"university\": 18\n",
    "    },\n",
    "    \"hobby\": [\"sports\", \"reading\"],\n",
    "    \"married\": false\n",
    "}\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我们可以用 `json.loads()` (load string) 方法从字符串中读取 `JSON` 数据："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{u'age': 24,\n",
      " u'ages for school': {u'high school': 15,\n",
      "                      u'middle school': 9,\n",
      "                      u'primary school': 6,\n",
      "                      u'university': 18},\n",
      " u'coding skills': [u'python',\n",
      "                    u'matlab',\n",
      "                    u'java',\n",
      "                    u'c',\n",
      "                    u'c++',\n",
      "                    u'ruby',\n",
      "                    u'scala'],\n",
      " u'hobby': [u'sports', u'reading'],\n",
      " u'married': False,\n",
      " u'name': u'echo'}\n"
     ]
    }
   ],
   "source": [
    "info = json.loads(info_string)\n",
    "\n",
    "pprint(info)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "此时，我们将原来的 `JSON` 数据变成了一个 `Python` 对象，在我们的例子中这个对象是个字典（也可能是别的类型，比如列表）："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(info)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "可以使用 `json.dumps()` 将一个 `Python` 对象变成 `JSON` 对象："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\"name\": \"echo\", \"age\": 24, \"married\": false, \"ages for school\": {\"middle school\": 9, \"university\": 18, \"high school\": 15, \"primary school\": 6}, \"coding skills\": [\"python\", \"matlab\", \"java\", \"c\", \"c++\", \"ruby\", \"scala\"], \"hobby\": [\"sports\", \"reading\"]}\n"
     ]
    }
   ],
   "source": [
    "info_json = json.dumps(info)\n",
    "\n",
    "print info_json"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "从中我们可以看到，生成的 `JSON` 字符串中，数组的元素顺序是不变的（始终是 `[\"python\", \"matlab\", \"java\", \"c\", \"c++\", \"ruby\", \"scala\"]`），而对象的元素顺序是不确定的。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 生成和读取 JSON 文件"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "与 `pickle` 类似，我们可以直接从文件中读取 `JSON` 数据，也可以将对象保存为 `JSON` 格式。\n",
    "\n",
    "- `json.dump(obj, file)` 将对象保存为 JSON 格式的文件\n",
    "- `json.load(file)` 从 JSON 文件中读取数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with open(\"info.json\", \"w\") as f:\n",
    "    json.dump(info, f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "可以查看 `info.json` 的内容："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\"name\": \"echo\", \"age\": 24, \"married\": false, \"ages for school\": {\"middle school\": 9, \"university\": 18, \"high school\": 15, \"primary school\": 6}, \"coding skills\": [\"python\", \"matlab\", \"java\", \"c\", \"c++\", \"ruby\", \"scala\"], \"hobby\": [\"sports\", \"reading\"]}\n"
     ]
    }
   ],
   "source": [
    "with open(\"info.json\") as f:\n",
    "    print f.read()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "从文件中读取数据："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{u'age': 24,\n",
      " u'ages for school': {u'high school': 15,\n",
      "                      u'middle school': 9,\n",
      "                      u'primary school': 6,\n",
      "                      u'university': 18},\n",
      " u'coding skills': [u'python',\n",
      "                    u'matlab',\n",
      "                    u'java',\n",
      "                    u'c',\n",
      "                    u'c++',\n",
      "                    u'ruby',\n",
      "                    u'scala'],\n",
      " u'hobby': [u'sports', u'reading'],\n",
      " u'married': False,\n",
      " u'name': u'echo'}\n"
     ]
    }
   ],
   "source": [
    "with open(\"info.json\") as f:\n",
    "    info_from_file = json.load(f)\n",
    "    \n",
    "pprint(info_from_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "删除生成的文件："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "os.remove(\"info.json\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}