Remove unreserved characters from the beginning of found URL

fixes #131
lipoja · Oct 21, 2022 · 1159482 · 1159482
1 parent 6bc2f30
commit 1159482
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 2 deletions.
diff --git a/tests/unit/test_find_urls.py b/tests/unit/test_find_urls.py
@@ -37,8 +37,8 @@
             ["https://example.com/what.com"],
         ),
         (
-            "https://i2.wp.com/siliconfilter.com/2011/06/example.jpg",
-            ["https://i2.wp.com/siliconfilter.com/2011/06/example.jpg"],
+            "* test link -https://www.example.com",
+            ["https://www.example.com"],
         ),
         (
             "https://www.test.org/paper/apostrophe'in-url",

diff --git a/urlextract/urlextract_core.py b/urlextract/urlextract_core.py
@@ -488,6 +488,9 @@ def _complete_url(
         # URL should not start with two backslashes
         if complete_url.startswith("//"):
             complete_url = complete_url[2:]
+        # URL should not start with unreserved characters
+        if complete_url.startswith(("-", ".", "~", "_")):
+            complete_url = complete_url[1:]
         if not self._is_domain_valid(
             complete_url, tld, check_dns=check_dns, with_schema_only=with_schema_only
         ):