Reddit EarthPorn automatic downloader

earthporn.py

import feedparser
import re
from bs4 import BeautifulSoup
import pprint
import os
import urllib.request
import imghdr
import sys
from time import localtime, strftime
from PIL import Image
import io

root = "F:/project/python/autobg/"

user_agent = "Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/37.0.2049.0 Safari/537.36"

earthporn = "http://www.reddit.com/r/EarthPorn/.rss"

rss = feedparser.parse(earthporn)

images = [ path.split(".")[0] for path in os.listdir(root + "images") ]

sys.stdout = open(root + "log.txt","a")

print(strftime("%Y-%m-%d %H:%M:%S", localtime()))

for entry in rss["entries"]:
	try:
		desc = BeautifulSoup(entry["description"])
		url = desc.find(text="[link]").parent["href"]
	
		if "/imgur.com/" in url:
			url = url.replace("/imgur.com/","/i.imgur.com/") + ".jpg"

		thumbnail = entry.media_thumbnail[0]["url"][-20:-4]
		
		if thumbnail in images:
			print("[Ignore]: Already exists")
			continue

		size = list(map(int,re.search("(\d+)[ Xx×]+(\d+)",entry["title"]).group(1,2)))
		if size[0]*size[1]/1366.0/768 < 0.8:
			print("[Ignore]: Image too small " + str(size))
			continue

		image = urllib.request.urlopen(urllib.request.Request(url, headers={'User-Agent': user_agent})).read()
		
		extension = Image.open(io.BytesIO(image)).format
		filename = thumbnail + "." + extension.lower()
		with open(root + "images/" + filename, "wb") as f:
			f.write(image)
		print("[Success]: " + filename + " " +str(size))

	except Exception as e:
		print("[Error]: " + str(e))