handle URLs without '/' after their TLD (#5252)

mikf · Feb 29, 2024 · 76581c1 · 76581c1
1 parent c006f9c
commit 76581c1
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 2 deletions.
diff --git a/gallery_dl/text.py b/gallery_dl/text.py
@@ -59,8 +59,14 @@ def ensure_http_scheme(url, scheme="https://"):
 def root_from_url(url, scheme="https://"):
     """Extract scheme and domain from a URL"""
     if not url.startswith(("https://", "http://")):
-        return scheme + url[:url.index("/")]
-    return url[:url.index("/", 8)]
+        try:
+            return scheme + url[:url.index("/")]
+        except ValueError:
+            return scheme + url
+    try:
+        return url[:url.index("/", 8)]
+    except ValueError:
+        return url
 
 
 def filename_from_url(url):

diff --git a/test/test_text.py b/test/test_text.py
@@ -121,12 +121,14 @@ def test_ensure_http_scheme(self, f=text.ensure_http_scheme):
 
     def test_root_from_url(self, f=text.root_from_url):
         result = "https://example.org"
+        self.assertEqual(f("https://example.org")     , result)
         self.assertEqual(f("https://example.org/")    , result)
         self.assertEqual(f("https://example.org/path"), result)
         self.assertEqual(f("example.org/")            , result)
         self.assertEqual(f("example.org/path/")       , result)
 
         result = "http://example.org"
+        self.assertEqual(f("http://example.org")      , result)
         self.assertEqual(f("http://example.org/")     , result)
         self.assertEqual(f("http://example.org/path/"), result)
         self.assertEqual(f("example.org/", "http://") , result)