Парсер ini-файлов - решение д/з для м/к по RegEx в Python

regex-hw.ipynb

{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import re\n",
      "import itertools as it"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = '''\n",
      "[sec 1] ; as]dasd\n",
      " k1    =v1; 123\n",
      " ;k3=\n",
      "\n",
      "[sec 2 ]\n",
      "a=b\n",
      "\n",
      "    [sec 1]\n",
      "k2= 123'''"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "res = {'sec 1': {'k1': 'v1', 'k2': 123}, 'sec 2 ': {'a': 'b'}}"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def test(fn):\n",
      "    return fn(data.split('\\n')) == res"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def load(lines):\n",
      "    re_section = re.compile(r'^\\s*\\[(\\w[\\w ]*)\\]\\s*(?:;.*)?$').match\n",
      "    re_pair = re.compile(r'^\\s*([\\w_]+)\\s*=\\s*(?:(\\d+)|(.+?))(?:;.*)?$').match\n",
      "    re_ignorable = re.compile(r'^(?:\\s*;.*)|(?:\\s*)$').match\n",
      "    \n",
      "    lines = it.ifilterfalse(re_ignorable, iter(lines))\n",
      "    res = {}\n",
      "    section = None\n",
      "    for idx, l in enumerate(lines, 1):\n",
      "        s = re_section(l)\n",
      "        if s:\n",
      "            section, = s.groups()\n",
      "            continue\n",
      "        p = re_pair(l)\n",
      "        if p:\n",
      "            if section is None:\n",
      "                raise ValueError('key-value pair ouside of section!')\n",
      "            k, i, v = p.groups()\n",
      "            res.setdefault(section, {})[k] = i and int(i) or v\n",
      "            continue\n",
      "        raise ValueError('line error: %3d: %r' % (idx, l))\n",
      "    return res\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print load(data.split('\\n'))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "{'sec 2 ': {'a': 'b'}, 'sec 1': {'k2': 123, 'k1': 'v1'}}\n"
       ]
      }
     ],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "test(load)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 7,
       "text": [
        "True"
       ]
      }
     ],
     "prompt_number": 7
    }
   ],
   "metadata": {}
  }
 ]
}