FIX: Fixing pre-commit check_links (#462)

.github/check_links.py

@@ -1,7 +1,9 @@
 import re
 import sys
 import os
+import time
 import requests
+from urllib.parse import urlsplit, urlunsplit
 from concurrent.futures import ThreadPoolExecutor, as_completed
 
 
@@ -20,17 +22,36 @@ def extract_urls(file_path):
         content = file.read()
     matches = URL_PATTERN.findall(content)
     # Flatten the list of tuples and filter out empty strings
-    urls = [url for match in matches for url in match if url]
+    urls = [strip_fragment(url) for match in matches for url in match if url]
     return urls
 
 
+def strip_fragment(url):
+    """
+    Removes the fragment (#...) from the URL, so the base URL can be checked.
+    """
+    parsed_url = urlsplit(url)
+    return urlunsplit((parsed_url.scheme, parsed_url.netloc, parsed_url.path, parsed_url.query, ""))
+
+
 def resolve_relative_url(base_path, url):
     if not url.startswith(("http://", "https://", "mailto:")):
         return os.path.abspath(os.path.join(os.path.dirname(base_path), url))
     return url
 
 
-def check_url(url):
+def check_url(url, retries=2, delay=2):
+    """
+    Check the validity of a URL, with retries if it fails.
+
+    Args:
+        url (str): URL to check.
+        retries (int, optional): Number of retries if the URL check fails. Defaults to 2.
+        delay (int, optional): Delay in seconds between retries. Defaults to 2.
+    Returns:
+        tuple: A tuple containing the URL and a boolean indicating whether it is valid.
+    """
+
     if (
         "http://localhost:" in url
         or url in skipped_urls
@@ -39,13 +60,23 @@ def check_url(url):
         or url.startswith("mailto:")
     ):
         return url, True
-    try:
-        response = requests.head(url, allow_redirects=True, timeout=5)
-        if response.status_code >= 400:
-            return url, False
-        return url, True
-    except requests.RequestException:
-        return url, False
+
+    attempts = 0
+    while attempts <= retries:
+        try:
+            response = requests.head(url, allow_redirects=True, timeout=5)
+            if response.status_code >= 400:
+                attempts += 1
+                if attempts > retries:
+                    return url, False
+                time.sleep(delay)
+            else:
+                return url, True
+        except requests.RequestException:
+            attempts += 1
+            if attempts > retries:
+                return url, False
+            time.sleep(delay)
 
 
 def check_links_in_file(file_path):