Tips/2016-03-21-Try-else.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Python 无处不在的 `else`\n",
    "\n",
    "我们都知道 Python 中 `else` 的基本用法是在条件控制语句中的 `if...elif...else...`，但是 `else` 还有两个其它的用途，一是用于循环的结尾，另一个是用在错误处理的 `try` 中。这原本是 Python 的标准语法，但由于和大部分其它编程语言的习惯不太一样，致使人们有意或无意地忽略了这些用法。另外，对于这些用法是否符合 [0x00 The Zen of Python](https://github.com/rainyear/pytips/blob/master/Tips/2016-03-06-The-Zen-of-Python.ipynb) 的原则以及该不该广泛使用也存在很多争议。例如在我看到的两本书里（[Effective Python](http://www.effectivepython.com/) VS [Write Idiomatic Python](https://jeffknupp.com/writing-idiomatic-python-ebook/)），两位作者就分别对其持有截然不同的态度。\n",
    "\n",
    "**循环中的 `else`**\n",
    "\n",
    "跟在循环后面的 `else` 语句只有在当循环内没出现 `break`，也就是正常循环完成时才会执行。首先我们来看一个插入排序法的例子："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[8, 12, 12, 34, 38, 68, 72, 78, 84, 90]\n"
     ]
    }
   ],
   "source": [
    "from random import randrange\n",
    "def insertion_sort(seq):\n",
    "    if len(seq) <= 1:\n",
    "        return seq\n",
    "    _sorted = seq[:1]\n",
    "    for i in seq[1:]:\n",
    "        inserted = False\n",
    "        for j in range(len(_sorted)):\n",
    "            if i < _sorted[j]:\n",
    "                _sorted = [*_sorted[:j], i, *_sorted[j:]]\n",
    "                inserted = True\n",
    "                break\n",
    "        if not inserted:\n",
    "            _sorted.append(i)\n",
    "    return _sorted\n",
    "\n",
    "print(insertion_sort([randrange(1, 100) for i in range(10)]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在这个例子中，对已排序的 `_sorted` 元素逐个与 `i` 进行比较，若 `i` 比已排序的所有元素都大，则只能排在已排序列表的最后。这时我们就需要一个额外的状态变量 `inserted` 来标记完成遍历循环还是中途被 `break`，在这种情况下，我们可以用 `else` 来取代这一状态变量："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 10, 27, 32, 32, 43, 50, 55, 80, 94]\n"
     ]
    }
   ],
   "source": [
    "def insertion_sort(seq):\n",
    "    if len(seq) <= 1:\n",
    "        return seq\n",
    "    _sorted = seq[:1]\n",
    "    for i in seq[1:]:\n",
    "        for j in range(len(_sorted)):\n",
    "            if i < _sorted[j]:\n",
    "                _sorted = [*_sorted[:j], i, *_sorted[j:]]\n",
    "                break\n",
    "        else:\n",
    "            _sorted.append(i)\n",
    "    return _sorted\n",
    "print(insertion_sort([randrange(1, 100) for i in range(10)]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我认为这是一个非常酷的做法！不过要注意的是，除了 `break` 可以触发后面的 `else` 语句，没有循环的时候也会："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loop failed!\n"
     ]
    }
   ],
   "source": [
    "while False:\n",
    "    print(\"Will never print!\")\n",
    "else:\n",
    "    print(\"Loop failed!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**错误捕捉中的 `else`**\n",
    "\n",
    "`try...except...else...finally` 流程控制语法用于捕捉可能出现的异常并进行相应的处理，其中 `except` 用于捕捉 `try` 语句中出现的错误；而 `else` 则用于处理**没有出现错误**的情况；`finally` 负责 `try` 语句的”善后工作“ ，无论如何都会执行。可以通过一个简单的例子来展示："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "result = 2.5\n",
      "divide finished!\n",
      "********************\n",
      "division by 0!\n",
      "divide finished!\n"
     ]
    }
   ],
   "source": [
    "def divide(x, y):\n",
    "    try:\n",
    "        result = x / y\n",
    "    except ZeroDivisionError:\n",
    "        print(\"division by 0!\")\n",
    "    else:\n",
    "        print(\"result = {}\".format(result))\n",
    "    finally:\n",
    "        print(\"divide finished!\")\n",
    "divide(5,2)\n",
    "print(\"*\"*20)\n",
    "divide(5,0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "当然，也可以用状态变量的做法来替代 `else`："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "result = 2.5\n",
      "divide finished!\n",
      "********************\n",
      "division by 0!\n",
      "divide finished!\n"
     ]
    }
   ],
   "source": [
    "def divide(x, y):\n",
    "    result = None\n",
    "    try:\n",
    "        result = x / y\n",
    "    except ZeroDivisionError:\n",
    "        print(\"division by 0!\")\n",
    "    if result is not None:\n",
    "        print(\"result = {}\".format(result))\n",
    "    print(\"divide finished!\")\n",
    "\n",
    "        \n",
    "divide(5,2)\n",
    "print(\"*\"*20)\n",
    "divide(5,0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**总结**\n",
    "\n",
    "有人觉得 `else` 的这些用法违反直觉或者是 **implicit** 而非 **explicit**，不值得提倡。但我觉得这种”判决“需要依赖具体的应用场景以及我们对 Python 的理解，并非一定要对新人友好的语法才算是 **explicit** 的。当然也不推荐在所有地方都使用这个语法，`for/while...else` 最大的缺点在于 `else` 是需要与 `for/file` 对齐的，如果是多层嵌套或者循环体太长的情况，就非常不适合用 `else`（回忆一下游标卡尺的梗就知道了:P）。只有在一些简短的循环控制语句中，我们通过 `else` 摆脱一些累赘的状态变量，这才是最 Pythonic 的应用场景！"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}