From ef1361ec39f30caff425bb91319182c887b073f8 Mon Sep 17 00:00:00 2001
From: Aroy-Art <Aroy-Art@pm.me>
Date: Tue, 4 Jun 2024 20:36:30 +0200
Subject: [PATCH] Add: util fuction to check if string is html

---
 archivist/utils/strings.py | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/archivist/utils/strings.py b/archivist/utils/strings.py
index 408e7af..027b881 100644
--- a/archivist/utils/strings.py
+++ b/archivist/utils/strings.py
@@ -1,6 +1,8 @@
 
 import math, re
 
+from bs4 import BeautifulSoup
+
 def get_urls(string:str = None) -> list:
     """
     Extracts URLs from a given string using regular expressions.
@@ -35,3 +37,17 @@ def convert_size(size_bytes: int) -> str:
         p = math.pow(1024, i)
         s = round(size_bytes / p, 2)
         return f"{s} {size_name[i]}"
+
+
+def is_string_html(string: str) -> bool:
+    '''
+    Check if string is HTML
+    
+    Parameters:
+        string (str): The string to be checked
+    
+    Returns:
+        bool: True if string is HTML, False otherwise
+    '''
+    return bool(BeautifulSoup(string, "html.parser").find())
+