Parse wiki markup with mwlib

parse_raw_wiki.py

from mwlib.refine import compat,core
from mwlib.parser import nodes
import pprint

def main():
  pp = pprint.PrettyPrinter(indent=2)

  raw = """
=Cake=
Modern cake, especially layer cakes, normally contain a
combination of [[flour]], [[sugar]], [[egg (food)|eggs]], and [[butter]].

==Varieties==
This is a subheading.

===Chocolate===
Oh and another nested subheading

=Related Cakes=

* Jaffa cakes!?
* This related section is not a subsection

Cool
"""

  # Returns raw 'token' classes that aren't much use
  #r = core.parse_txt(raw)

  ### This is the magic
  # Returns nicer classes, Paragraph, PreFormatted etc.
  parsed = compat.parse_txt(raw)
  
  ### Everything below here is just as an example

  f = open('test.txt', 'w')

  for section in parsed:
    recursive_parse(f, section, 0)

  f.close


def recursive_parse(f, node, indent):
  tabs    = indent * "  "
  indent += 1
  print tabs + node.__class__.__name__

  if isinstance(node, nodes.Section):
    level = u"=" * node.level
    title = level + node.children[0].asText()
    title = title.rstrip()
    f.write(title)

    # Skip the text child, move onto the rest
    for child in node.children[1:]:
      recursive_parse(f, child, indent)
    return

  elif isinstance(node, nodes.Table):
    # Don't process children
    print tabs + "Table: Skipping"
    return

  elif isinstance(node, nodes.Text):
    txt = node.asText()
    if txt != "\n": # and re.match("^\s+$", txt) == None:

      txt = txt.rstrip('\n')
      #re.sub("\n", '', txt)

      #f.write('<text>' + txt.encode('UTF-8') + '</text>')
      f.write(txt.encode('UTF-8'))

  else:
    f.write("???")

  if node.children is not None:
    for child in node.children:
      recursive_parse(f, child, indent)

if __name__ == '__main__':
  main()