Parse HTML document, upload local images to Imgur and update document with imgur link

parse-html-upload-imgur.py

import os
import sys
import json
import requests
from base64 import b64encode
from bs4 import BeautifulSoup
from urllib.parse import urlparse

IMGUR_CLIENT_ID = ""

def _soup(file):
  with open(file, "r") as f:
    soup = BeautifulSoup(f.read(), 'html.parser')
    return soup

url = "https://api.imgur.com/3/upload.json"

def _upload_imgur(name, data, client_id=IMGUR_CLIENT_ID):
  file = b64encode(data)

  headers = {"Authorization": f"Client-ID {client_id}"}

  req = requests.post(
      url, 
      headers = headers,
      data = {
          'image': file,
          'type': 'base64',
          'name': name,
      }
  )

  data = req.json()
  return data["data"]["link"]

def _upload_image_local(path, client_id=IMGUR_CLIENT_ID):
  with open(path, "rb") as img:
    print(f"[+] uploading {path}")
    return _upload_imgur(os.path.basename(path), img.read(), client_id)

def _upload_image_url(url, client_id=IMGUR_CLIENT_ID):
  print(f"[+] fetching from url {url}")
  resp = requests.get(url)
  print(f"[+] uploading {url}")
  return _upload_imgur(os.path.basename(url), resp.content, client_id)

def uri_validator(x):
    try:
        result = urlparse(x)
        return all([result.scheme, result.netloc])
    except:
        return False

if __name__ == "__main__":
  if len(sys.argv) != 2:
    print("""Usage: python upload.py <file.html>
file.html can contain local files or remote resources, these will be uploaded to imgur""")
    exit(1)
  
  file = sys.argv[1]

  soup = _soup(file)

  images = soup.find_all("img")

  for img in images:
    src = img["src"]

    try:
      link = ""
      
      if uri_validator(src):
        link = _upload_image_url(src)
      else:
        link = _upload_image_local(src)
      
      print(f"[+] uploaded {src} to {link}")

      img["src"] = link
    except Exception as e:
      print(f"[!] failed to upload {src}: {e}")

  hrefs = soup.find_all("a")

  for href in hrefs:
    lnk = href["href"]
    print(f"[+] changing {lnk} to {{.URL}}")
    href["href"] = "{{.URL}}"

  with open(sys.argv[1], "wb") as f_output:
    f_output.write(soup.prettify("utf-8"))