photos_sample.py

from selenium import webdriver
from selenium.webdriver.common.keys import Keys
import time
import pickle

# Setup the browser. Could be headless, just testing.
options = webdriver.ChromeOptions()
options.add_argument('--ignore-certificate-errors')
options.add_argument("--test-type")
options.binary_location = "/usr/bin/chromium"
driver = webdriver.Chrome(chrome_options=options)

# Get the initial page
driver.get('https://photos.google.com')
# Open already generated cookies to avoid loging in
cookies = pickle.load(open("cookies.pkl", "rb"))
for cookie in cookies:
     driver.add_cookie(cookie)

# Search for a specific file by name
driver.get('https://photos.google.com/search/IMG_20191008_183025')
# This is too hacky, there must be a better way to tell apart the image elements.
elems = driver.find_elements_by_xpath('//a[@tabindex=0][@aria-label]')
# Also too hacky, if there is only one result, it is the last element with this selector.
elems[len(elems)-1].click()

elem = driver.switch_to.active_element
elem.send_keys(Keys.SHIFT, 'd')