Add --exclude-caches option, for excluding CACHEDIR.TAG directories

See http://www.brynosaurus.com/cachedir/spec.html

attic/archiver.py

@@ -17,7 +17,8 @@ from attic.key import key_creator
 from attic.helpers import Error, location_validator, format_time, \
     format_file_mode, ExcludePattern, exclude_path, adjust_patterns, to_localtime, \
     get_cache_dir, get_keys_dir, format_timedelta, prune_within, prune_split, \
-    Manifest, remove_surrogates, update_excludes, format_archive, check_extension_modules, Statistics
+    Manifest, remove_surrogates, update_excludes, format_archive, check_extension_modules, Statistics, \
+    is_cachedir
 from attic.remote import RepositoryServer, RemoteRepository
 
 
@@ -124,7 +125,7 @@ Type "Yes I am sure" if you understand this and want to continue.\n""")
                     continue
             else:
                 restrict_dev = None
-            self._process(archive, cache, args.excludes, skip_inodes, path, restrict_dev)
+            self._process(archive, cache, args.excludes, args.exclude_caches, skip_inodes, path, restrict_dev)
         archive.save()
         if args.stats:
             t = datetime.now()
@@ -140,7 +141,7 @@ Type "Yes I am sure" if you understand this and want to continue.\n""")
             print('-' * 78)
         return self.exit_code
 
-    def _process(self, archive, cache, excludes, skip_inodes, path, restrict_dev):
+    def _process(self, archive, cache, excludes, exclude_caches, skip_inodes, path, restrict_dev):
         if exclude_path(path, excludes):
             return
         try:
@@ -163,6 +164,8 @@ Type "Yes I am sure" if you understand this and want to continue.\n""")
             except IOError as e:
                 self.print_error('%s: %s', path, e)
         elif stat.S_ISDIR(st.st_mode):
+            if exclude_caches and is_cachedir(path):
+                return
             archive.process_item(path, st)
             try:
                 entries = os.listdir(path)
@@ -170,7 +173,7 @@ Type "Yes I am sure" if you understand this and want to continue.\n""")
                 self.print_error('%s: %s', path, e)
             else:
                 for filename in sorted(entries):
-                    self._process(archive, cache, excludes, skip_inodes,
+                    self._process(archive, cache, excludes, exclude_caches, skip_inodes,
                                   os.path.join(path, filename), restrict_dev)
         elif stat.S_ISLNK(st.st_mode):
             archive.process_symlink(path, st)
@@ -519,6 +522,9 @@ Type "Yes I am sure" if you understand this and want to continue.\n""")
         subparser.add_argument('--exclude-from', dest='exclude_files',
                                type=argparse.FileType('r'), action='append',
                                metavar='EXCLUDEFILE', help='read exclude patterns from EXCLUDEFILE, one per line')
+        subparser.add_argument('--exclude-caches', dest='exclude_caches',
+                               action='store_true', default=False,
+                               help='exclude directories that contain a CACHEDIR.TAG file (http://www.brynosaurus.com/cachedir/spec.html)')
         subparser.add_argument('-c', '--checkpoint-interval', dest='checkpoint_interval',
                                type=int, default=300, metavar='SECONDS',
                                help='write checkpoint every SECONDS seconds (Default: 300)')

attic/helpers.py

@@ -245,6 +245,25 @@ class ExcludePattern(IncludePattern):
     def __repr__(self):
         return '%s(%s)' % (type(self), self.pattern)
 
+def is_cachedir(path):
+    """Determines whether the specified path is a cache directory (and
+    therefore should potentially be excluded from the backup) according to
+    the CACHEDIR.TAG protocol
+    (http://www.brynosaurus.com/cachedir/spec.html).
+    """
+
+    tag_filename = 'CACHEDIR.TAG'
+    tag_contents = b'Signature: 8a477f597d28d172789f06886806bc55'
+    tag_path = os.path.join(path, 'CACHEDIR.TAG')
+    try:
+        if os.path.exists(tag_path):
+            with open(tag_path, 'rb') as tag_file:
+                tag_data = tag_file.read(len(tag_contents))
+                if tag_data == tag_contents:
+                    return True
+    except OSError as e:
+        raise
+    return False
 
 def walk_path(path, skip_inodes=None):
     st = os.lstat(path)

attic/testsuite/archiver.py

@@ -106,12 +106,14 @@ class ArchiverTestCaseBase(AtticTestCase):
 
 class ArchiverTestCase(ArchiverTestCaseBase):
 
-    def create_regular_file(self, name, size=0):
+    def create_regular_file(self, name, size=0, contents=None):
         filename = os.path.join(self.input_path, name)
         if not os.path.exists(os.path.dirname(filename)):
             os.makedirs(os.path.dirname(filename))
         with open(filename, 'wb') as fd:
-            fd.write(b'X' * size)
+            if contents is None:
+                contents = b'X' * size
+            fd.write(contents)
 
     def create_test_files(self):
         """Create a minimal test case including all supported file types
@@ -178,6 +180,17 @@ class ArchiverTestCase(ArchiverTestCaseBase):
             self.attic('extract', '--exclude-from=' + self.exclude_file_path, self.repository_location + '::test')
         self.assert_equal(sorted(os.listdir('output/input')), ['file1', 'file3'])
 
+    def test_exclude_caches(self):
+        self.attic('init', self.repository_location)
+        self.create_regular_file('file1', size=1024 * 80)
+        self.create_regular_file('cache1/CACHEDIR.TAG', contents = b'Signature: 8a477f597d28d172789f06886806bc55 extra stuff')
+        self.create_regular_file('cache2/CACHEDIR.TAG', contents = b'invalid signature')
+        self.attic('create', '--exclude-caches', self.repository_location + '::test', 'input')
+        with changedir('output'):
+            self.attic('extract', self.repository_location + '::test')
+        self.assert_equal(sorted(os.listdir('output/input')), ['cache2', 'file1'])
+        self.assert_equal(sorted(os.listdir('output/input/cache2')), ['CACHEDIR.TAG'])
+
     def test_path_normalization(self):
         self.attic('init', self.repository_location)
         self.create_regular_file('dir1/dir2/file', size=1024 * 80)