A script that morphologically parses a specified URL

Experiment to try morphological analysis for each specified URL. I tried to remove HTML tags with a regular expression, but I can't remove them.

urlmecab.py

#!/user/bin/env python
# -*- coding: utf-8 -*-
import urllib
import sys
import MeCab
import re


while True:
	search_url = raw_input(u"input URL: ")
	

	def Mecab_file():	
		req = urllib.urlopen(search_url)
		dlText = req.read()

		mt = MeCab.Tagger("mecabrc")
		data = []
		p = re.compile(r"<[^>]*?>")
		sus = p.sub("", dlText)
		data.append(sus)


		node = mt.parseToNode("\n".join(data))
		words = {}
		
		while node:
			word = node.surface
			if word and node.posid >=36 and node.posid <=67:
				if not words.has_key(word):
					words[word] = 0
				words[word] += 1
			node = node.next
		word_items = words.items()
		word_items.sort()
		word_items.reverse()
		for word, count in word_items:
			print word, count
			
	if search_url:
		Mecab_file()
	else:
		break

Extract only nouns with the part of speech ID of MeCab.

if word and node.posid >=36 and node.posid <=67:

If you change this part, you may be able to play a lot. Loop as long as you keep typing the URL. Loop break with blank enter. http://〜入力する必要あり。