from html to jupyter notebook

1from bs4 import BeautifulSoup
2import json
3import urllib.request
4url = 'http://nbviewer.jupyter.org/url/jakevdp.github.com/downloads/notebooks/XKCD_plots.ipynb'
5response = urllib.request.urlopen(url)
6#  for local html file
7# response = open("/Users/note/jupyter/notebook.html")
8text = response.read()
9
10soup = BeautifulSoup(text, 'lxml')
11# see some of the html
12print(soup.div)
13dictionary = {'nbformat': 4, 'nbformat_minor': 1, 'cells': [], 'metadata': {}}
14for d in soup.findAll("div"):
15    if 'class' in d.attrs.keys():
16        for clas in d.attrs["class"]:
17            if clas in ["text_cell_render", "input_area"]:
18                # code cell
19                if clas == "input_area":
20                    cell = {}
21                    cell['metadata'] = {}
22                    cell['outputs'] = []
23                    cell['source'] = [d.get_text()]
24                    cell['execution_count'] = None
25                    cell['cell_type'] = 'code'
26                    dictionary['cells'].append(cell)
27
28                else:
29                    cell = {}
30                    cell['metadata'] = {}
31
32                    cell['source'] = [d.decode_contents()]
33                    cell['cell_type'] = 'markdown'
34                    dictionary['cells'].append(cell)
35open('notebook.ipynb', 'w').write(json.dumps(dictionary))
36

1from IPython.display import HTML
2HTML(filename='myhtml.html')
3