Read URLs from XML Sitemap

read_xml_sitemaps.py

sitemaps = {}

for (sitemap_url, lasmod) in sitemap_index.items():
  if(sitemap_url.find("post") > 0):
    print(sitemap_url)
    
    if 1: # for testing
      r = requests.get(sitemap_url)
      xml = r.text

      soup = BeautifulSoup(xml)
      URLTags = soup.find_all("url")

      print("The number of URLs are {0}".format(len(URLTags)))

      for URL in URLTags:
          sitemaps[URL.findNext("loc").text] = URL.findNext("lastmod").text

      #print(xml_sitemap) #for testing
  
  #break #for testing