#! /usr/bin/env python

import sys
reload(sys)
sys.setdefaultencoding('utf8')

from lxml import html

def my_print(item, search):
	if search.split('/')[-1] == 'a':
		print item.xpath('@href')[0], '|', item.text.strip()
	else:
		print item


search = sys.argv[1]

page_list = sys.argv[2:]
for page in page_list:
	orig = open(page).read()
	t = html.fromstring(orig)

	list = t.xpath(search)
	if isinstance(list, basestring):
		my_print(list, search)
	else:
		for item in list:
			my_print(item, search)