| """Functions related to processing images""" | |
| import io | |
| import time | |
| import warnings | |
| from typing import Union | |
| import numpy as np | |
| import requests | |
| from PIL import Image | |
| from wrapt_timeout_decorator.wrapt_timeout_decorator import timeout | |
| def download_image_with_timeout(url: str) -> np.ndarray: | |
| while True: | |
| # Get the data from the URL, and try again if it fails | |
| response = requests.get(url) | |
| if response.status_code != 200: | |
| time.sleep(1) | |
| continue | |
| # Convert the data to a NumPy array | |
| byte_file = io.BytesIO(response.content) | |
| image = np.asarray(Image.open(byte_file), dtype=np.uint8) | |
| return image | |
| def process_image_url(url: str) -> Union[None, dict]: | |
| """Process the URL and extract the article. | |
| Args: | |
| url (str): The URL. | |
| Returns: | |
| dict or None: | |
| The processed article, or None if the URL could not be parsed. | |
| """ | |
| # Ignore warnings while processing images | |
| with warnings.catch_warnings(): | |
| warnings.simplefilter("ignore") | |
| try: | |
| image = download_image_with_timeout(url) | |
| except: # noqa | |
| return None | |
| if image is None: | |
| return None | |
| else: | |
| try: | |
| return dict( | |
| url=url, pixels=image, height=image.shape[0], width=image.shape[1] | |
| ) | |
| except IndexError: | |
| return None | |