Work around Python 2.6+ urljoin bug.

doc/changelog.txt

@@ -9,6 +9,9 @@ Fixes:
   Closes: SF bug #3035754
 - file: Prevent truncation of UNC paths on Windows systems.
   Closes: SF bug #3017391
+- url: Work around a Python bug cutting off characters when joining an
+  URL that starts with semicolon.
+  Closes: SF bug #3056136
 
 Changes:
 - checking: Caches are now size-restricted to limit the memory

linkcheck/checker/urlbase.py

@@ -52,6 +52,10 @@ def urljoin (parent, url, scheme):
     """
     if url.startswith(scheme+":"):
         return url
+    # work around a Python 2.6/3.1 bug cutting off characters when the URL
+    # begins with semicolon
+    if url.startswith(';'):
+        url = "./%s" % url
     return urlparse.urljoin(parent, url)
 
 

tests/test_urlbuild.py

@@ -22,6 +22,7 @@ import unittest
 import linkcheck.configuration
 import linkcheck.director
 import linkcheck.checker.httpurl
+import linkcheck.checker.urlbase
 
 
 def get_test_aggregate ():
@@ -47,3 +48,10 @@ class TestUrlBuild (unittest.TestCase):
                aggregate, parent_url=parent_url)
         o.build_url()
         self.assertEqual(o.url, 'http://')
+
+    def test_urljoin (self):
+        parent_url = "http://localhost:8001/test"
+        base_url = ";param=value"
+        scheme = 'http'
+        res = linkcheck.checker.urlbase.urljoin(parent_url, base_url, scheme)
+        self.assertEqual(res, 'http://localhost:8001/;param=value')