Fix long-standing bugs with MANIFEST.in parsing on Windows (#6884).

These regex changes fix a number of issues for distutils on Windows:
- #6884: impossible to include a file starting with 'build'
- #9691 and #14004: sdist includes too many files
- #13193: test_filelist failures

This commit replaces the incorrect changes done in 0a94e2f807c7 and
90b30d62caf2 to fix #13193; we were too eager to fix the test failures
and I did not study the code enough before greenlighting patches.  This
time we have unit tests from the problems reported by users to be sure
we have the right fix.

Thanks to Nadeem Vawda for his help.

1 files changed, 100 insertions, 30 deletions

diff --git a/Lib/distutils/tests/test_filelist.py b/Lib/distutils/tests/test_filelist.py
index f3304b2e33..e82bc3d2ed 100644
--- a/Lib/distutils/tests/test_filelist.py
+++ b/Lib/distutils/tests/test_filelist.py
@@ -1,4 +1,5 @@
 """Tests for distutils.filelist."""
+import os
 import re
 import unittest
 from distutils import debug
@@ -9,6 +10,26 @@ from distutils.filelist import glob_to_re, translate_pattern, FileList
 from test.support import captured_stdout, run_unittest
 from distutils.tests import support
 
+MANIFEST_IN = """\
+include ok
+include xo
+exclude xo
+include foo.tmp
+include buildout.cfg
+global-include *.x
+global-include *.txt
+global-exclude *.tmp
+recursive-include f *.oo
+recursive-exclude global *.x
+graft dir
+prune dir3
+"""
+
+
+def make_local_path(s):
+    """Converts '/' in a string to os.sep"""
+    return s.replace('/', os.sep)
+
 
 class FileListTestCase(support.LoggingSilencer,
                        unittest.TestCase):
@@ -22,16 +43,62 @@ class FileListTestCase(support.LoggingSilencer,
         self.clear_logs()
 
     def test_glob_to_re(self):
-        # simple cases
-        self.assertEqual(glob_to_re('foo*'), 'foo[^/]*\\Z(?ms)')
-        self.assertEqual(glob_to_re('foo?'), 'foo[^/]\\Z(?ms)')
-        self.assertEqual(glob_to_re('foo??'), 'foo[^/][^/]\\Z(?ms)')
-
-        # special cases
-        self.assertEqual(glob_to_re(r'foo\\*'), r'foo\\\\[^/]*\Z(?ms)')
-        self.assertEqual(glob_to_re(r'foo\\\*'), r'foo\\\\\\[^/]*\Z(?ms)')
-        self.assertEqual(glob_to_re('foo????'), r'foo[^/][^/][^/][^/]\Z(?ms)')
-        self.assertEqual(glob_to_re(r'foo\\??'), r'foo\\\\[^/][^/]\Z(?ms)')
+        sep = os.sep
+        if os.sep == '\\':
+            sep = re.escape(os.sep)
+
+        for glob, regex in (
+            # simple cases
+            ('foo*', r'foo[^%(sep)s]*\Z(?ms)'),
+            ('foo?', r'foo[^%(sep)s]\Z(?ms)'),
+            ('foo??', r'foo[^%(sep)s][^%(sep)s]\Z(?ms)'),
+            # special cases
+            (r'foo\\*', r'foo\\\\[^%(sep)s]*\Z(?ms)'),
+            (r'foo\\\*', r'foo\\\\\\[^%(sep)s]*\Z(?ms)'),
+            ('foo????', r'foo[^%(sep)s][^%(sep)s][^%(sep)s][^%(sep)s]\Z(?ms)'),
+            (r'foo\\??', r'foo\\\\[^%(sep)s][^%(sep)s]\Z(?ms)')):
+            regex = regex % {'sep': sep}
+            self.assertEqual(glob_to_re(glob), regex)
+
+    def test_process_template_line(self):
+        # testing  all MANIFEST.in template patterns
+        file_list = FileList()
+        l = make_local_path
+
+        # simulated file list
+        file_list.allfiles = ['foo.tmp', 'ok', 'xo', 'four.txt',
+                              'buildout.cfg',
+                              # filelist does not filter out VCS directories,
+                              # it's sdist that does
+                              l('.hg/last-message.txt'),
+                              l('global/one.txt'),
+                              l('global/two.txt'),
+                              l('global/files.x'),
+                              l('global/here.tmp'),
+                              l('f/o/f.oo'),
+                              l('dir/graft-one'),
+                              l('dir/dir2/graft2'),
+                              l('dir3/ok'),
+                              l('dir3/sub/ok.txt'),
+                             ]
+
+        for line in MANIFEST_IN.split('\n'):
+            if line.strip() == '':
+                continue
+            file_list.process_template_line(line)
+
+        wanted = ['ok',
+                  'buildout.cfg',
+                  'four.txt',
+                  l('.hg/last-message.txt'),
+                  l('global/one.txt'),
+                  l('global/two.txt'),
+                  l('f/o/f.oo'),
+                  l('dir/graft-one'),
+                  l('dir/dir2/graft2'),
+                 ]
+
+        self.assertEqual(file_list.files, wanted)
 
     def test_debug_print(self):
         file_list = FileList()
@@ -117,6 +184,7 @@ class FileListTestCase(support.LoggingSilencer,
         self.assertEqual(file_list.allfiles, ['a.py', 'b.txt'])
 
     def test_process_template(self):
+        l = make_local_path
         # invalid lines
         file_list = FileList()
         for action in ('include', 'exclude', 'global-include',
@@ -127,7 +195,7 @@ class FileListTestCase(support.LoggingSilencer,
 
         # include
         file_list = FileList()
-        file_list.set_allfiles(['a.py', 'b.txt', 'd/c.py'])
+        file_list.set_allfiles(['a.py', 'b.txt', l('d/c.py')])
 
         file_list.process_template_line('include *.py')
         self.assertEqual(file_list.files, ['a.py'])
@@ -139,31 +207,31 @@ class FileListTestCase(support.LoggingSilencer,
 
         # exclude
         file_list = FileList()
-        file_list.files = ['a.py', 'b.txt', 'd/c.py']
+        file_list.files = ['a.py', 'b.txt', l('d/c.py')]
 
         file_list.process_template_line('exclude *.py')
-        self.assertEqual(file_list.files, ['b.txt', 'd/c.py'])
+        self.assertEqual(file_list.files, ['b.txt', l('d/c.py')])
         self.assertNoWarnings()
 
         file_list.process_template_line('exclude *.rb')
-        self.assertEqual(file_list.files, ['b.txt', 'd/c.py'])
+        self.assertEqual(file_list.files, ['b.txt', l('d/c.py')])
         self.assertWarnings()
 
         # global-include
         file_list = FileList()
-        file_list.set_allfiles(['a.py', 'b.txt', 'd/c.py'])
+        file_list.set_allfiles(['a.py', 'b.txt', l('d/c.py')])
 
         file_list.process_template_line('global-include *.py')
-        self.assertEqual(file_list.files, ['a.py', 'd/c.py'])
+        self.assertEqual(file_list.files, ['a.py', l('d/c.py')])
         self.assertNoWarnings()
 
         file_list.process_template_line('global-include *.rb')
-        self.assertEqual(file_list.files, ['a.py', 'd/c.py'])
+        self.assertEqual(file_list.files, ['a.py', l('d/c.py')])
         self.assertWarnings()
 
         # global-exclude
         file_list = FileList()
-        file_list.files = ['a.py', 'b.txt', 'd/c.py']
+        file_list.files = ['a.py', 'b.txt', l('d/c.py')]
 
         file_list.process_template_line('global-exclude *.py')
         self.assertEqual(file_list.files, ['b.txt'])
@@ -175,50 +243,52 @@ class FileListTestCase(support.LoggingSilencer,
 
         # recursive-include
         file_list = FileList()
-        file_list.set_allfiles(['a.py', 'd/b.py', 'd/c.txt', 'd/d/e.py'])
+        file_list.set_allfiles(['a.py', l('d/b.py'), l('d/c.txt'),
+                                l('d/d/e.py')])
 
         file_list.process_template_line('recursive-include d *.py')
-        self.assertEqual(file_list.files, ['d/b.py', 'd/d/e.py'])
+        self.assertEqual(file_list.files, [l('d/b.py'), l('d/d/e.py')])
         self.assertNoWarnings()
 
         file_list.process_template_line('recursive-include e *.py')
-        self.assertEqual(file_list.files, ['d/b.py', 'd/d/e.py'])
+        self.assertEqual(file_list.files, [l('d/b.py'), l('d/d/e.py')])
         self.assertWarnings()
 
         # recursive-exclude
         file_list = FileList()
-        file_list.files = ['a.py', 'd/b.py', 'd/c.txt', 'd/d/e.py']
+        file_list.files = ['a.py', l('d/b.py'), l('d/c.txt'), l('d/d/e.py')]
 
         file_list.process_template_line('recursive-exclude d *.py')
-        self.assertEqual(file_list.files, ['a.py', 'd/c.txt'])
+        self.assertEqual(file_list.files, ['a.py', l('d/c.txt')])
         self.assertNoWarnings()
 
         file_list.process_template_line('recursive-exclude e *.py')
-        self.assertEqual(file_list.files, ['a.py', 'd/c.txt'])
+        self.assertEqual(file_list.files, ['a.py', l('d/c.txt')])
         self.assertWarnings()
 
         # graft
         file_list = FileList()
-        file_list.set_allfiles(['a.py', 'd/b.py', 'd/d/e.py', 'f/f.py'])
+        file_list.set_allfiles(['a.py', l('d/b.py'), l('d/d/e.py'),
+                                l('f/f.py')])
 
         file_list.process_template_line('graft d')
-        self.assertEqual(file_list.files, ['d/b.py', 'd/d/e.py'])
+        self.assertEqual(file_list.files, [l('d/b.py'), l('d/d/e.py')])
         self.assertNoWarnings()
 
         file_list.process_template_line('graft e')
-        self.assertEqual(file_list.files, ['d/b.py', 'd/d/e.py'])
+        self.assertEqual(file_list.files, [l('d/b.py'), l('d/d/e.py')])
         self.assertWarnings()
 
         # prune
         file_list = FileList()
-        file_list.files = ['a.py', 'd/b.py', 'd/d/e.py', 'f/f.py']
+        file_list.files = ['a.py', l('d/b.py'), l('d/d/e.py'), l('f/f.py')]
 
         file_list.process_template_line('prune d')
-        self.assertEqual(file_list.files, ['a.py', 'f/f.py'])
+        self.assertEqual(file_list.files, ['a.py', l('f/f.py')])
         self.assertNoWarnings()
 
         file_list.process_template_line('prune e')
-        self.assertEqual(file_list.files, ['a.py', 'f/f.py'])
+        self.assertEqual(file_list.files, ['a.py', l('f/f.py')])
         self.assertWarnings()