bpo-36035: fix Path.rglob for broken links (GH-11988)

jstucke · pitrou · commit d5c120f7eb6f · 2019-05-21T19:44:40.000+02:00
Links creating an infinite symlink loop would raise an exception.
diff --git a/Lib/pathlib.py b/Lib/pathlib.py
@@ -7,7 +7,7 @@
 import re
 import sys
 from _collections_abc import Sequence
-from errno import EINVAL, ENOENT, ENOTDIR, EBADF
+from errno import EINVAL, ENOENT, ENOTDIR, EBADF, ELOOP
 from operator import attrgetter
 from stat import S_ISDIR, S_ISLNK, S_ISREG, S_ISSOCK, S_ISBLK, S_ISCHR, S_ISFIFO
 from urllib.parse import quote_from_bytes as urlquote_from_bytes
@@ -35,10 +35,11 @@
 #
 
 # EBADF - guard against macOS `stat` throwing EBADF
-_IGNORED_ERROS = (ENOENT, ENOTDIR, EBADF)
+_IGNORED_ERROS = (ENOENT, ENOTDIR, EBADF, ELOOP)
 
 _IGNORED_WINERRORS = (
     21,  # ERROR_NOT_READY - drive exists but is not accessible
+    1921,  # ERROR_CANT_RESOLVE_FILENAME - fix for broken symlink pointing to itself
 )
 
 def _ignore_error(exception):
@@ -520,7 +521,13 @@ def _select_from(self, parent_path, is_dir, exists, scandir):
             cf = parent_path._flavour.casefold
             entries = list(scandir(parent_path))
             for entry in entries:
-                if not self.dironly or entry.is_dir():
+                entry_is_dir = False
+                try:
+                    entry_is_dir = entry.is_dir()
+                except OSError as e:
+                    if not _ignore_error(e):
+                        raise
+                if not self.dironly or entry_is_dir:
                     name = entry.name
                     casefolded = cf(name)
                     if self.pat.match(casefolded):
diff --git a/Lib/test/test_pathlib.py b/Lib/test/test_pathlib.py
@@ -1221,7 +1221,8 @@ class _BasePathTest(object):
     #  |-- dirE  # No permissions
     #  |-- fileA
     #  |-- linkA -> fileA
-    #  `-- linkB -> dirB
+    #  |-- linkB -> dirB
+    #  `-- brokenLinkLoop -> brokenLinkLoop
     #
 
     def setUp(self):
@@ -1252,6 +1253,8 @@ def cleanup():
             self.dirlink(os.path.join('..', 'dirB'), join('dirA', 'linkC'))
             # This one goes upwards, creating a loop.
             self.dirlink(os.path.join('..', 'dirB'), join('dirB', 'linkD'))
+            # Broken symlink (pointing to itself).
+            os.symlink('brokenLinkLoop',  join('brokenLinkLoop'))
 
     if os.name == 'nt':
         # Workaround for http://bugs.python.org/issue13772.
@@ -1384,7 +1387,7 @@ def test_iterdir(self):
         paths = set(it)
         expected = ['dirA', 'dirB', 'dirC', 'dirE', 'fileA']
         if support.can_symlink():
-            expected += ['linkA', 'linkB', 'brokenLink']
+            expected += ['linkA', 'linkB', 'brokenLink', 'brokenLinkLoop']
         self.assertEqual(paths, { P(BASE, q) for q in expected })
 
     @support.skip_unless_symlink
@@ -1465,6 +1468,7 @@ def test_rglob_symlink_loop(self):
                   'fileA',
                   'linkA',
                   'linkB',
+                  'brokenLinkLoop',
                   }
         self.assertEqual(given, {p / x for x in expect})
 
diff --git a/Misc/NEWS.d/next/Core and Builtins/2019-02-22-14-30-19.bpo-36035.-6dy1y.rst b/Misc/NEWS.d/next/Core and Builtins/2019-02-22-14-30-19.bpo-36035.-6dy1y.rst
@@ -0,0 +1 @@
+Added fix for broken symlinks in combination with pathlib

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	`+Added fix for broken symlinks in combination with pathlib`