fork download 
  1. import json
  2.  
  3. import lxml.html
  4.  
  5.  
  6. def first(iterable, default=None):
  7.     try:
  8.         return next(iter(iterable))
  9.     except StopIteration:
  10.         return None
  11.  
  12.  
  13. def get_tree(node):
  14.     childs = []
  15.     for child_node in node.iterchildren():
  16.         child = get_child(child_node)
  17.         childs.append(child)
  18.     return childs
  19.  
  20.  
  21. def get_child(child_node):
  22.     child = {}
  23.     tag = child_node.tag
  24.     if tag == 'ul':
  25.         child['subtree'] = get_tree(child_node)
  26.     elif tag == 'li':
  27.         child['title'] = child_node.find('span').text_content()
  28.         child['link'] = first(child_node.xpath('span/a/@href'))
  29.         ul = child_node.find('ul')
  30.         if ul is not None:
  31.             child['subtree'] = get_tree(ul)
  32.     else:
  33.         raise Exception("Unexpected tag: {}".format(tag))
  34.     return child
  35.  
  36.  
  37. def main():
  38.     with open('./rutracker_cr_forum_map.html') as fin:
  39.         html = lxml.html.fromstring(fin.read())
  40.     fmap = html.get_element_by_id('f-map')
  41.     tree = get_tree(fmap)
  42.     with open('rutracker_cr_forum_map.json', 'w') as fout:
  43.         json.dump(tree, fout, ensure_ascii=False, indent=4)
  44.  
  45.  
  46. if __name__ == "__main__":
  47.     main()
  48.  
Runtime error #stdin #stdout #stderr 0.04s 98240KB
comments (?)
stdin 
Standard input is empty
stdout 
Standard output is empty
stderr 
Traceback (most recent call last):
  File "./prog.py", line 47, in <module>
    main()
  File "./prog.py", line 38, in main
    with open('./rutracker_cr_forum_map.html') as fin:
FileNotFoundError: [Errno 2] No such file or directory: './rutracker_cr_forum_map.html'
https://ideone.com/MZ8flY
language:
Python 3 nbc (python 3.7.3)
created:
8 years ago
visibility:
public

Share or Embed source code

Discover > Sphere Engine API

The brand new service which powers Ideone!

Discover > IDE Widget

Widget for compiling and running the source code in a web browser!