fork download
copy 
  1. import json
  2.  
  3. import attr
  4. import lxml.html
  5.  
  6.  
  7. @attr.s
  8. class Child:
  9.     title = attr.ib(default=None)
  10.     link = attr.ib(default=None)
  11.     subtree = attr.ib(default=[])
  12.  
  13.  
  14. def first(iterable, default=None):
  15.     try:
  16.         return next(iter(iterable))
  17.     except StopIteration:
  18.         return None
  19.  
  20.  
  21. def get_tree(node):
  22.     childs = []
  23.     for child_node in node.iterchildren():
  24.         child = get_child(child_node)
  25.         childs.append(child)
  26.     return childs
  27.  
  28.  
  29. def get_child(child_node):
  30.     child = Child()
  31.     tag = child_node.tag
  32.     if tag == 'ul':
  33.         child.subtree = get_tree(child_node)
  34.     elif tag == 'li':
  35.         child.title = child_node.find('span').text_content()
  36.         child.link = first(child_node.xpath('span/a/@href'))
  37.         ul = child_node.find('ul')
  38.         if ul is not None:
  39.             child.subtree = get_tree(ul)
  40.     else:
  41.         raise Exception("Unexpected tag: {}".format(tag))
  42.     return attr.asdict(child)
  43.  
  44.  
  45. def main():
  46.     with open('./rutracker_cr_forum_map.html') as fin:
  47.         html = lxml.html.fromstring(fin.read())
  48.     fmap = html.get_element_by_id('f-map')
  49.     tree = get_tree(fmap)
  50.     with open('rutracker_cr_forum_map.json', 'w') as fout:
  51.         json.dump(tree, fout, ensure_ascii=False, indent=4)
  52.  
  53.  
  54. if __name__ == "__main__":
  55.     main()
  56.  
Runtime error #stdin #stdout #stderr 0.02s 30744KB
comments (?)
stdin
copy 
Standard input is empty
stdout
copy 
Standard output is empty
stderr
copy 
Traceback (most recent call last):
  File "./prog.py", line 3, in <module>
    import attr
ImportError: No module named 'attr'
https://ideone.com/16nOrh
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!