#!/usr/bin/python

"""
    This script will scrape the r-project.org machine learning selection and
    format the packages in github markdown style for this
    awesome-machine-learning repo.
"""

from pyquery import PyQuery as pq
import urllib
import codecs
import random

text_file = codecs.open("Packages.txt", encoding='utf-8', mode="w")
d = pq(url='http://cran.r-project.org/web/views/MachineLearning.html',
       opener=lambda url, **kw: urllib.urlopen(url).read())

for e in d("li").items():
    package_name = e("a").html()
    package_link = e("a")[0].attrib['href']
    if '..' in package_link:
        package_link = package_link.replace("..",
                                            'http://cran.r-project.org/web')
        dd = pq(url=package_link, opener=lambda url,
                **kw: urllib.urlopen(url).read())
        package_description = dd("h2").html()
        text_file.write(" [%s](%s) - %s \n" % (package_name, package_link,
                                               package_description))
        # print("* [%s](%s) - %s" % (package_name,package_link,
        #                            package_description))