Add: util func to get all urls in a string

archivist/utils/strings.py

@@ -0,0 +1,15 @@
+
+
+import re
+
+def get_urls(string):
+    '''
+    A function that returns all URLs from a string.
+    '''
+    regex = re.compile(
+        r'((([A-Za-z]{3,9}:(?:\/\/)?)(?:[-;:&=\+\$,\w]+@)?[A-Za-z0-9.-]+|(?:www.|[-;:&=\+\$,\w]+@)[A-Za-z0-9.-]+)((?:\/[\+~%\/.\w_]*)?\??(?:[-\+=&;%@.\w_]*)#?(?:[.\!\/\\w]*))?)',
+        re.IGNORECASE)
+
+    return [x[0] for x in re.findall(regex, string)]
+
+