{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make a new `get_XXX_url`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "```python\n",
    "@staticmethod\n",
    "def get_soup_url(url):\n",
    "    return url\n",
    "\n",
    "@staticmethod\n",
    "def get_pdf_url(url):\n",
    "    return url\n",
    "\n",
    "@staticmethod\n",
    "def get_tex_url(url):\n",
    "    return url\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[32m[success]\u001b[0m local driver can be built.\n",
      "\u001b[31m[failure]\u001b[0m remote driver can't be built.\n",
      "DRIVER_TYPE: \u001b[32mlocal\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "import sys\n",
    "sys.path.append(\"../tests/\")\n",
    "import inspect\n",
    "from data import JournalData\n",
    "from gummy import journals\n",
    "from gummy.journals import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Genetics\n"
     ]
    }
   ],
   "source": [
    "journal = input().lower()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Current Method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "crawler = journals.get(journal)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    @staticmethod\n",
      "    def get_soup_url(url):\n",
      "        return url\n",
      "\n",
      "    @staticmethod\n",
      "    def get_pdf_url(url):\n",
      "        return url\n",
      "\n",
      "    @staticmethod\n",
      "    def get_tex_url(url):\n",
      "        return url\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(inspect.getsource(crawler.get_soup_url))\n",
    "print(inspect.getsource(crawler.get_pdf_url))\n",
    "print(inspect.getsource(crawler.get_tex_url))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visit web page and Collect various url"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "https://www.genetics.org/content/176/4/2177\n"
     ]
    }
   ],
   "source": [
    "url = JournalData.get(journal)\n",
    "print(url)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "urls = [\n",
    "    \"https://www.genetics.org/content/176/4/2177\",\n",
    "    \"https://www.genetics.org/content/genetics/176/4/2177.full.pdf\",\n",
    "    \"https://www.genetics.org/content/176/4/2177.full.pdf\",\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create URL conversion function (staticmethod)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### `get_soup_url`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "https://www.genetics.org/content/176/4/2177\n",
      "https://www.genetics.org/content/176/4/2177\n",
      "https://www.genetics.org/content/176/4/2177\n"
     ]
    }
   ],
   "source": [
    "for url in urls:\n",
    "    print(url.rstrip(\".full.pdf\").replace(\"/content/genetics/\", \"/content/\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### `get_pdf_url`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "https://www.genetics.org/content/genetics/176/4/2177.full.pdf\n",
      "https://www.genetics.org/content/genetics/176/4/2177.full.pdf\n",
      "https://www.genetics.org/content/genetics/176/4/2177.full.pdf\n"
     ]
    }
   ],
   "source": [
    "for url in urls:\n",
    "    print(url.rstrip(\".full.pdf\").replace(\"/content/genetics/\", \"/content/\").replace(\"/content/\", \"/content/genetics/\")+\".full.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### `get_tex_url`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Created Conversion Method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    @staticmethod\n",
      "    def get_soup_url(url):\n",
      "        return url.rstrip(\".full.pdf\").replace(\"/content/genetics/\", \"/content/\")\n",
      "\n",
      "    @staticmethod\n",
      "    def get_pdf_url(url):\n",
      "        return GeneticsCrawler.get_soup_url(url).replace(\"/content/\", \"/content/genetics/\")+\".full.pdf\"  \n",
      "\n",
      "    @staticmethod\n",
      "    def get_tex_url(url):\n",
      "        return url\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(inspect.getsource(crawler.get_soup_url))\n",
    "print(inspect.getsource(crawler.get_pdf_url))\n",
    "print(inspect.getsource(crawler.get_tex_url))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "hide_input": false,
  "kernelspec": {
   "display_name": "Python 3 (/usr/local/bin/)",
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}