CodeLyokoScrapping.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:5b4d434866d32c721d4cc0943f1a59bccbc1de2609d4e2cd1162b860c87f8856"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from bs4 import BeautifulSoup\n",
      "from requests import get"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "BASE_URL = 'http://www.episodesdecodelyoko.com'"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def retrieve_ids(saison):\n",
      "    url = BASE_URL + '/' + 'saison.php?saison=' + str(saison)\n",
      "    soup = BeautifulSoup(get(url).content)\n",
      "    return map(\n",
      "               lambda _: re.sub(r'.*id=(\\d+).*', r'\\1', _),\n",
      "               filter(\n",
      "                    lambda _: _.find('lecteur=li')>-1,\n",
      "                    filter(None,\n",
      "                           map(lambda _:_.get('href'),\n",
      "                               soup.findAll('a')\n",
      "                           )\n",
      "                    )\n",
      "                )\n",
      "            )"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 169
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def get_mp4_url(idx):\n",
      "    soup = BeautifulSoup(get(BASE_URL+'/episode.php?id='+str(idx)+'&lecteur=li').content)\n",
      "    return list(map(lambda _: re.sub(r\"file: '([^']+)'.*\", r'\\1', _),\n",
      "             filter(lambda _: _.startswith('file:'),\n",
      "                    map(lambda _: _.strip(),\n",
      "                        list(filter(lambda _: _.find('jwplayer(')>-1,\n",
      "                                    map(lambda _: _.text, soup.find_all('script'))\n",
      "                             )\n",
      "                        )[0].splitlines()\n",
      "                    )\n",
      "            )\n",
      "        )\n",
      "    )[0]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 191
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "urls = []\n",
      "for s in range(5):\n",
      "    for idx in retrieve_ids(s):\n",
      "        urls.append(get_mp4_url(idx))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 193
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "with open('url_code_lyoko.txt', 'w') as f:\n",
      "    for u in urls: f.write(u+'\\n')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 194
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}