Based on Richard idea I decided to loop on xPath values and not link text(which can be the same) the code I ended up with is:
import lxml.html as lh
import urllib2
from selenium import webdriver
htmlObject = lh.parse(urllib2.urlopen(start_url))
listOfPaths = htmlObject.xpath('//ul[@class="directory dir-col"]/li/a')
listOfLinkPathes = []
for link in listOfPaths:
listOfLinkPathes.append(htmlObject.getpath(link))
for linkPath in listOfLinkPathes:
mydriver.find_element_by_xpath(linkPath).click()
mydriver.back()