From 3f0c2fae5e18c448bd1791920500c976d44fc321 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Adri=C3=A1n=20Chaves?= <adrian@chaves.io>
Date: Wed, 12 Apr 2023 09:28:28 +0200
Subject: [PATCH] Skip test_skip_bad_links based on the w3lib version

---
 scrapy/linkextractors/lxmlhtml.py |  6 +++---
 tests/test_linkextractors.py      | 16 +++++++++++-----
 tox.ini                           |  1 +
 3 files changed, 15 insertions(+), 8 deletions(-)

diff --git a/scrapy/linkextractors/lxmlhtml.py b/scrapy/linkextractors/lxmlhtml.py
index 0d1b76531..23cbd0116 100644
--- a/scrapy/linkextractors/lxmlhtml.py
+++ b/scrapy/linkextractors/lxmlhtml.py
@@ -1,8 +1,8 @@
 """
 Link extractor based on lxml.html
 """
-import operator
 import logging
+import operator
 from functools import partial
 from urllib.parse import urljoin, urlparse
 
@@ -94,8 +94,8 @@ class LxmlParserLinkExtractor:
             try:
                 url = safe_url_string(url, encoding=response_encoding)
             except ValueError:
-                logger.debug(f"Skipping extraction of bad link {url}")
-                continue  # Disregard badly formatted urls
+                logger.debug(f"Skipping extraction of link with bad URL {url!r}")
+                continue
 
             # to fix relative links after process_value
             url = urljoin(response_url, url)
diff --git a/tests/test_linkextractors.py b/tests/test_linkextractors.py
index 784fdb658..e1ec19601 100644
--- a/tests/test_linkextractors.py
+++ b/tests/test_linkextractors.py
@@ -1,7 +1,10 @@
 import pickle
 import re
 import unittest
-import sys 
+
+from packaging.version import Version
+from pytest import mark
+from w3lib import __version__ as w3lib_version
 
 from scrapy.http import HtmlResponse, XmlResponse
 from scrapy.link import Link
@@ -817,13 +820,16 @@ class LxmlLinkExtractorTestCase(Base.LinkExtractorTestCase):
     def test_restrict_xpaths_with_html_entities(self):
         super().test_restrict_xpaths_with_html_entities()
 
-    @unittest.skipIf(
-        sys.version_info < (3, 8),
-        reason="some library for python 3.7 so is less strict so bad links like htis don't crash scrapy",
+    @mark.skipif(
+        Version(w3lib_version) < Version("2.0.0"),
+        reason=(
+            "Before w3lib 2.0.0, w3lib.url.safe_url_string would not complain "
+            "about an invalid port value."
+        ),
     )
     def test_skip_bad_links(self):
         html = b"""
-        <a href="http://Some wierd html : http://example.com/like_this">Why would you do this?</a>
+        <a href="http://example.org:non-port">Why would you do this?</a>
         <a href="http://example.org/item2.html">Good Link</a>
         <a href="http://example.org/item3.html">Good Link 2</a>
         """
diff --git a/tox.ini b/tox.ini
index 5a9d9cf29..873e7662b 100644
--- a/tox.ini
+++ b/tox.ini
@@ -101,6 +101,7 @@ install_command =
     python -I -m pip install {opts} {packages}
 
 [testenv:pinned]
+basepython = python3.7
 deps =
     {[pinned]deps}
     PyDispatcher==2.0.5