#!/usr/bin/env python
# coding: utf-8

# # Making a calendar from a Python release-schedule PEP
# 
# Python 3.6 has its release schedule posted in [PEP 494](https://www.python.org/dev/peps/pep-0494/).
# 
# I'm interested in having this information into my calendar.
# Rather than the sensible thing of manually adding a few events to my calendar,
# I'm going to build a calendar automatically in a notebook.
# 
# **Why?** Because notebooks are fun,
# and I've been using [soupy](https://soupy.readthedocs.io)
# to practice some functional programming.
# 
# I'm going to use [soupy](https://soupy.readthedocs.io) to extract the information,
# and [icalendar](https://icalendar.readthedocs.io) to build the calendar data.

# In[1]:


import requests
from soupy import Soupy, Q


# Fetch the page:

# In[2]:


url = 'https://www.python.org/dev/peps/pep-0494/'
soup = Soupy(requests.get(url).text, 'html.parser')


# The schedule is helpfully encapsulated in `div#schedule`:

# In[3]:


print(soup.find(id='schedule').val())


# We can see that each entry is a pretty simple `<li>` tag,
# with a release and date separated by a colon.
# 
# We can use Soupy's functional style to quickly parse this information.
# We want to do:
# 
# - find `#schedule`
# - for each `li` in `#schedule`:
#   - get the text before and after the `':'`

# In[4]:


raw_data = soup \
    .find(id='schedule') \
    .find_all('li') \
    .each(
        Q.text.strip().split(':')
    )
raw_data.val()


# Let's parse those dates before we get ahead of ourselves:

# In[5]:


import datetime
import re

date_pat = re.compile(r'\d{4}-\d{2}-\d{2}')

def parse_date(title_date):
    title, date_string = title_date
    date_part = date_pat.search(date_string).group()
    date = datetime.datetime.strptime(date_part, '%Y-%m-%d').date()
    return (title, date)
    
    
data = raw_data.each(Q.map(parse_date))
data.val()


# Now that we have the data, we can build the calendar with [icalendar](https://icalendar.readthedocs.io).
# 
# We want each release to be an all-day event,
# and have a summary like "Python release 3.6.0 beta 1".
# For good behavior reasons, we will give each item a UID,
# so that importing the same event multiple times doesn't create duplicates.

# In[6]:


import hashlib
import icalendar

cal = icalendar.Calendar()

def add_event(cal, title, date):
    evt = icalendar.Event()
    # add Python to the summary
    summary = 'Python ' + title
    evt.add('summary', 'Python ' + title)
    evt.add('dtstart', date)
    evt.add('dtend', date)
    # give it a UID for stability on repeated imports
    key = b'python-release-%s' % title.encode('utf8')
    evt.add('uid', hashlib.md5(key).hexdigest())
    cal.add_component(evt)

data.each(Q.map(lambda title_date: add_event(cal, *title_date)))
print(cal.to_ical().decode())


# Now we can save this file to disk:

# In[7]:


with open('python36.ics', 'wb') as f:
    f.write(cal.to_ical())


# and import the resulting `python36.ics` into your calendar application of choice.
# 
# You can see the result of importing this file into Google Calendar:
# 

# In[8]:


get_ipython().run_cell_magic('html', '', '<iframe src="https://calendar.google.com/calendar/embed?src=arcfq33k7n1docj2bpbda8jhg0%40group.calendar.google.com" style="border: 0" width="800" height="600" frameborder="0" scrolling="no"></iframe>\n')


# <center>
# This notebook is by Min RK and placed in the Public Domain (or Creative Commons CC0, if you prefer).
# </center>