Fixed matching Windows paths.

CHANGES.rst

@@ -2,6 +2,12 @@
 Change History
 ==============
 
+0.3.2 (TBD)
+-----------
+
+- Issue #6: Fixed matching Windows paths.
+
+
 0.3.1 (2014-09-17)
 ------------------
 

pathspec/__init__.py

@@ -1,4 +1,4 @@
-# coding: utf-8
+# encoding: utf-8
 """
 The *pathspec* package provides pattern matching for file paths. So far
 this only includes gitignore style pattern matching.
@@ -18,8 +18,8 @@
 __license__ = "MPL 2.0"
 __project__ = "pathspec"
 __status__ = "Development"
-__updated__ = "2014-09-17"
-__version__ = "0.3.1"
+__updated__ = "2014-11-07"
+__version__ = "0.3.2b1"
 
 from .gitignore import GitIgnorePattern
 from .pathspec import PathSpec

pathspec/compat.py

@@ -1,12 +1,23 @@
-# coding: utf-8
+# encoding: utf-8
 """
 This module provides compatibility between Python 2 and 3. Hardly
 anything is used by this project to constitute including `six`_.
 
 .. _`six`: http://pythonhosted.org/six
 """
 
-try:
+import sys
+
+if sys.version_info[0] < 3:
+	# Python 2.
 	string_types = (basestring,)
-except NameError:
+
+	def viewkeys(mapping):
+		return mapping.viewkeys()
+
+else:
+	# Python 3.
 	string_types = (str,)
+
+	def viewkeys(mapping):
+		return mapping.keys()

pathspec/gitignore.py

@@ -1,4 +1,4 @@
-# coding: utf-8
+# encoding: utf-8
 """
 This module implements gitignore style pattern matching which
 incorporates POSIX glob patterns.

pathspec/pathspec.py

@@ -1,4 +1,4 @@
-# coding: utf-8
+# encoding: utf-8
 """
 This module provides an object oriented interface for pattern matching
 of files.
@@ -7,7 +7,7 @@
 import collections
 
 from . import util
-from .compat import string_types
+from .compat import string_types, viewkeys
 
 
 class PathSpec(object):
@@ -63,16 +63,27 @@ def from_lines(cls, pattern_factory, lines):
 		lines = [pattern_factory(line) for line in lines if line]
 		return cls(lines)
 
-	def match_files(self, files):
+	def match_files(self, files, separators=None):
 		"""
 		Matches the files to this path-spec.
 
 		*files* (``Iterable`` of ``str``) contains the files to be matched
 		against *patterns*.
 
-		Returns the matched files (``set`` of ``str``).
+		*separators* (``Container`` of ``str``) optionally contains the path
+		separators to normalize. This does not need to include the POSIX
+		path separator (`/`), but including it will not affect the results.
+		Default is ``None`` to determine the separators based upon the
+		current operating system by examining `os.sep` and `os.altsep`. To
+		prevent normalization, pass an empty container (e.g., an empty tuple
+		`()`).
+
+		Returns the matched files (``Iterable`` of ``str``).
 		"""
-		return util.match_files(self.patterns, files)
+		file_map = util.normalize_files(files, separators=separators)
+		matched_files = util.match_files(self.patterns, viewkeys(file_map))
+		for path in matched_files:
+			yield file_map[path]
 
 	def match_tree(self, root):
 		"""
@@ -81,7 +92,7 @@ def match_tree(self, root):
 
 		*root* (``str``) is the root directory to search for files.
 
-		Returns the matched files (``set`` of ``str``).
+		Returns the matched files (``Iterable`` of ``str``).
 		"""
 		files = util.iter_tree(root)
 		return self.match_files(files)

pathspec/pattern.py

@@ -1,4 +1,4 @@
-# coding: utf-8
+# encoding: utf-8
 """
 This module provides the base definition for patterns.
 """

pathspec/tests/__init__.py

@@ -0,0 +1,7 @@
+# encoding: utf-8
+"""
+This package provides tests.
+"""
+
+from .test_pathspec import PathSpecTest
+from .test_gitignore import GitIgnoreTest

pathspec/tests/test_gitignore.py

@@ -1,4 +1,4 @@
-# coding: utf-8
+# encoding: utf-8
 """
 This script tests ``GitIgnorePattern``.
 """

pathspec/tests/test_pathspec.py

@@ -0,0 +1,35 @@
+# encoding: utf-8
+"""
+This script tests ``PathSpec``.
+"""
+
+import unittest
+
+import pathspec
+
+class PathSpecTest(unittest.TestCase):
+	"""
+	The ``PathSpecTest`` class tests the ``PathSpec`` class.
+	"""
+
+	def test_01_windows_paths(self):
+		"""
+		Tests that Windows paths will be properly normalized and matched.
+		"""
+		spec = pathspec.PathSpec.from_lines('gitignore', [
+			'*.txt',
+			'!test1/',
+		])
+		results = set(spec.match_files([
+			'src\\test1\\a.txt',
+			'src\\test1\\b.txt',
+			'src\\test1\\c\\c.txt',
+			'src\\test2\\a.txt',
+			'src\\test2\\b.txt',
+			'src\\test2\\c\\c.txt',
+		], separators=('\\',)))
+		self.assertEquals(results, {
+			'src\\test2\\a.txt',
+			'src\\test2\\b.txt',
+			'src\\test2\\c\\c.txt',
+		})

pathspec/util.py

@@ -1,4 +1,4 @@
-# coding: utf-8
+# encoding: utf-8
 """
 This module provides utility methods for dealing with path-specs.
 """
@@ -9,6 +9,13 @@
 
 from .compat import string_types
 
+NORMALIZE_PATH_SEPS = [sep for sep in [os.sep, os.altsep] if sep and sep != '/']
+"""
+*NORMALIZE_PATH_SEPS* (``list`` of ``str``) contains the path separators
+that need to be normalized to the POSIX separator for the current
+operating system.
+"""
+
 _registered_patterns = {}
 """
 *_registered_patterns* (``dict``) maps a name (``str``) to the
@@ -66,8 +73,8 @@ def match_files(patterns, files):
 	*patterns* (``Iterable`` of ``pathspec.Pattern``) contains the
 	patterns to use.
 
-	*files* (``Iterable`` of ``str``) contains the files to be matched
-	against *patterns*.
+	*files* (``Iterable`` of ``str``) contains the normalized files to be
+	matched against *patterns*.
 
 	Returns the matched files (``set`` of ``str``).
 	"""
@@ -82,6 +89,29 @@ def match_files(patterns, files):
 				return_files.difference_update(result_files)
 	return return_files
 
+def normalize_files(files, separators=None):
+	"""
+	Normalizes the file paths to use the POSIX path separator (i.e., `/`).
+
+	*files* (``Iterable`` of ``str``) contains the file paths to be
+	normalized.
+
+	*separators* (``Container`` of ``str``) optionally contains the path
+	separators to normalize.
+
+	Returns a ``dict`` mapping the normalized file path (``str``) to the
+	original file path (``str``)
+	"""
+	if separators is None:
+		separators = NORMALIZE_PATH_SEPS
+	file_map = {}
+	for path in files:
+		norm = path
+		for sep in separators:
+			norm = norm.replace(sep, '/')
+		file_map[norm] = path
+	return file_map
+
 def register_pattern(name, pattern_factory, override=None):
 	"""
 	Registers the specified pattern factory.

setup.py

@@ -1,4 +1,4 @@
-# coding: utf-8
+# encoding: utf-8
 
 import io
 from setuptools import setup, find_packages
@@ -33,5 +33,5 @@
 	],
 	license=__license__,
 	packages=find_packages(),
-	test_suite='pathspec.tests.test_gitignore.GitIgnoreTest',
+	test_suite='pathspec.tests',
 )