mirror of
https://github.com/bellingcat/auto-archiver.git
synced 2026-06-13 05:38:29 +03:00
35 lines
1.4 KiB
Python
35 lines
1.4 KiB
Python
from loguru import logger
|
|
import time, uuid, os
|
|
from selenium.common.exceptions import TimeoutException
|
|
|
|
from . import Enricher
|
|
from ..utils import Webdriver
|
|
from ..core import Media, Metadata
|
|
|
|
class ScreenshotEnricher(Enricher):
|
|
name = "screenshot_enricher"
|
|
|
|
@staticmethod
|
|
def configs() -> dict:
|
|
return {
|
|
"width": {"default": 1280, "help": "width of the screenshots"},
|
|
"height": {"default": 720, "help": "height of the screenshots"},
|
|
"timeout": {"default": 60, "help": "timeout for taking the screenshot"}
|
|
}
|
|
|
|
def enrich(self, to_enrich: Metadata) -> None:
|
|
url = to_enrich.get_url()
|
|
logger.debug(f"Enriching screenshot for {url=}")
|
|
with Webdriver(self.width, self.height, self.timeout, 'facebook.com' in url) as driver:
|
|
try:
|
|
driver.get(url)
|
|
time.sleep(2)
|
|
screenshot_file = os.path.join(to_enrich.get_tmp_dir(), f"screenshot_{str(uuid.uuid4())[0:8]}.png")
|
|
driver.save_screenshot(screenshot_file)
|
|
to_enrich.add_media(Media(filename=screenshot_file), id="screenshot")
|
|
except TimeoutException:
|
|
logger.info("TimeoutException loading page for screenshot")
|
|
except Exception as e:
|
|
logger.error(f"Got error while loading webdriver for screenshot enricher: {e}")
|
|
# return None
|