{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Web scraping avec Python\n",
    "## RoboBrowser\n",
    "\n",
    "- RoboBrowser est une librairie python qui permet de simuler un comportement  de navigation dans un navigateur web.\n",
    "- Elle permet aussi d'extraire des éléments précis d'information d'une page et de les structurer. \n",
    "- C'est la plus populaire qui combine les avantages de `beautifulsoup` avec la flexibilité de `requests`. \n",
    "- Elle est disponible pour Python 2.7 et Python 3.6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "from robobrowser import RoboBrowser"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On crée un objet navigateur et on navigue vers une première URL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "browser = RoboBrowser(history=True,parser=\"lxml\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "browser.open('https://meteo.gc.ca/')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ouverture d'un formulaire à partir d'une de ses propriétés (ici, son identifiant unique)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "form = browser.get_form(id=\"cityjump\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En affichant le formulaire, on peut voir les différents champs disponibles, ainsi que les valeurs par défauts qui sont attribuées, si applicable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<RoboForm city=, lang=f, unit=>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "form"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En utilisant la propriété `value`, on peut remplir le formulaire"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "form['city'].value = 'Québec'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On envoie ensuite le formulaire, comme si on cliquait sur le bouton d'envoi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "browser.submit_form(form)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'dl.mrgn-bttm-0'"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"dl.mrgn-bttm-0\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On obtient le résultat de la page suivante dans le navigatuer. À partir de ce résultat, on peut extraire différents éléments en utilisant le sélecteur `CSS`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "location_xml = browser. \\\n",
    "select('.col-sm-10')[0]. \\\n",
    "select('dl.mrgn-bttm-0')[0]. \\\n",
    "select('dd.mrgn-bttm-0')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lorsqu'un motif se répète, on préfère alors créer une fonction qui va extraire l'élément selon des paramètres. La librairie RoboBrowser ne gere pas les noeuds enfants de la structure XML (sélecteur `nth_child()`), mais on peut utiliser les listes de Python pour répliquer un comportement similaire."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"div.div-column:nth-child(4) > div:nth-child(2) > p:nth-child(2) > span:nth-child(1)\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def temperature_xml(jour): \n",
    "    return browser. \\\n",
    "    select('div.div-column')[jour]. \\\n",
    "    select('div')[1]. \\\n",
    "    select('p.mrgn-bttm-0')[0]. \\\n",
    "    select('span')[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On peut assembler les données extraites dans un dictionnaire python et les utiliser dans son application."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'city': 'Aéroport int. Lesage de Québec',\n",
       " 'date': '13h00 HNE le samedi 9 décembre 2017',\n",
       " 'temperature': ['-1°C\\n',\n",
       "  '-2°C\\n',\n",
       "  '-13°C\\n',\n",
       "  '-4°C\\n',\n",
       "  '-8°C\\n',\n",
       "  '-11°C\\n',\n",
       "  '-11°C\\n']}"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "{'city': location_xml[0].text, \n",
    " 'date': location_xml[1].text,\n",
    " 'temperature': [temperature_xml(i).text for i in range(0,7)]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}