Webpage Thumbnails In Python

thumb.py

#!/usr/bin/python
# -*- coding: utf-8 -*-

from cgi import parse_qs, escape
from selenium import webdriver
from uuid import uuid4
import cStringIO
import Image
import json

basePath = "/var/www/example.com/path/to/thumb/"
baseUrl = "http://example.com/path/to/thumbs/"
screenSize = (1024, 768)

def application(environ, start_response):

    d = parse_qs(environ['QUERY_STRING'])
    
    if not all(k in d for k in ['url', 'width', 'height']):
        output = json.dumps({'success': False})
    else:
        url = d['url'][0]
        width = int(d['width'][0])
        height = int(d['height'][0])
        output = ""
        
        driver = webdriver.PhantomJS(service_log_path=basePath + "ghostdriver.log")
        driver.set_script_timeout(30)
        driver.set_window_size(screenSize[0], screenSize[1])
        driver.get(url)
        
        screenshot_fname = str(uuid4()) + ".png"
        screenshot_fpath = "{0}thumbs/{1}".format(basePath, screenshot_fname)
        screenshot = cStringIO.StringIO(driver.get_screenshot_as_png())
        screenshot.seek(0) 
        with open(screenshot_fpath, 'w+') as sfile:
            image = Image.open(screenshot)
            image.thumbnail((width, height), Image.ANTIALIAS)
            image.save(screenshot_fpath, 'PNG')
        screenshot.close()
        output = json.dumps({'success': True, 'thumb': baseUrl + screenshot_fname})
    
    status = '200 OK'
    response_headers = [('Content-type', 'application/json'),
                        ('Content-Length', str(len(output)))]
    start_response(status, response_headers)
    return [output]

WebpageThumbnailsInPython.md

#Webpage Thumbnails In Python A small WSGI script that uses selenium's headless PhantomJS driver to capture webpage screenshots, and PIL to resize them. Cropping options could easily be added.

##Dependencies

• Python selenium
• Python PIL
• PhantomJS