Handle special filesystem entries explicitly (#11)

* Handle special filesystem entries explicitly

- Added fs.entry_kind(): classifies via lstat() mode bits alone
  ("file", "dir", "symlink", "fifo", "socket", "block_device",
  "char_device", "unknown") - never opens the entry, so classifying a
  FIFO can't block even without a writer.
- rsync() now classifies every source entry before touching it and
  yields a new Actions.EXCLUDE for anything outside file/dir/symlink,
  instead of falling through to copy_direntry()'s file branch, which
  would either block forever opening a writerless FIFO for read, or
  raise an OSError that initiate_backup() treats as fatal (aborting the
  whole backup for one socket).
- initiate_backup() records exclusions in the completion manifest
  (keyed by snapshot-relative path, valued with the excluded kind), and
  no longer nest_hardlinks or counts a recurring exclusion as a change
  worth keeping a new snapshot for.
- Chose the "recorded exclusion" policy the plan allows, rather than
  fail-fast: a FIFO or socket living in a backed-up tree is not itself
  a backup failure, and previously turned into one.

Note: rsync_ext() (external rsync) is unaffected by this change and
keeps whatever its own --archive flags do with special files; the P1
item on Python/external implementation parity covers reconciling that.
The pre-existing test_src_is_socket failure (AF_UNIX path length on
macOS tmp_path) is untouched here - it's separately tracked under P1
test fixes - but this PR adds working socket/FIFO/char-device coverage
via a short-path tempdir.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

* Address Copilot review: fix stale exclude, exclusion-only diffs, test portability

- fs.rsync(): remove the existing dst_entry before yielding EXCLUDE for
  a src entry with an unsupported kind, so an entry inherited via
  hardlink from a prior snapshot doesn't linger in the new one.
- backup.initiate_backup(): compare this run's excluded set against
  the previous snapshot's manifest; a change in exclusions alone now
  counts as backup_changed, so the updated manifest gets written.
- tests: skip mkfifo/AF_UNIX-dependent tests on platforms without
  support, and use TemporaryDirectory/context managers instead of
  manual mkdtemp/unlink for more robust cleanup.

* Remove unused shutil import from test_rsync.py

Left over after switching the socket/fifo tests to
tempfile.TemporaryDirectory(), per Copilot's suppressed comment.

* Fix rm_direntry() no-op on special dst entries

rm_direntry() only handled file/symlink/dir - a FIFO, socket, or
device node at dst (e.g. inherited from an external-rsync snapshot)
silently survived removal, since is_file/is_dir/is_symlink all
return False for it. Any non-directory entry is unlinkable the same
way, so invert the branching: recurse for directories, unlink
everything else.

---------

Co-authored-by: Maxim Shulga <maxim.shulga@alludo.com>
Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>

tests/test_integration.py

@@ -1081,3 +1081,35 @@ def test_external_rsync_snapshot_still_gets_checksums(integration_dirs):
     manifest = bk._read_manifest(bk._get_backup_marker(snapshot).path)
     assert manifest["checksums"]
     assert bk.verify_snapshot(snapshot) == []
+
+
+@pytest.mark.skipif(not hasattr(os, "mkfifo"),
+                    reason="os.mkfifo not available on this platform")
+def test_fifo_excluded_and_recorded_without_blocking(integration_dirs):
+    """
+    A FIFO in the source must not block the backup (no writer is ever
+    attached), must not abort it, and must be recorded in the manifest
+    as an exclusion rather than silently disappearing.
+    """
+    backups_dir, source_dir = integration_dirs
+    (source_dir / "file1.txt").write_text("content1")
+    os.mkfifo(str(source_dir / "a_fifo"))
+
+    snapshot = bk.initiate_backup(
+        sources=[str(source_dir)],
+        backups_dir=str(backups_dir),
+        dry_run=False,
+    )
+
+    assert snapshot is not None
+    source_name = os.path.basename(str(source_dir))
+    manifest = bk._read_manifest(bk._get_backup_marker(snapshot).path)
+    assert manifest["excluded"] == {
+        os.path.join(source_name, "a_fifo"): "fifo"
+    }
+    assert not os.path.lexists(
+        os.path.join(snapshot.path, source_name, "a_fifo")
+    )
+    assert os.path.exists(
+        os.path.join(snapshot.path, source_name, "file1.txt")
+    )

tests/test_rsync.py

@@ -709,3 +709,103 @@ class TestRsyncHardlinkImmutability:
 
         assert os.lstat(str(external_target)).st_mode == external_mode_before
         assert (external_target / "canary").read_text() == "do-not-touch"
+
+
+class TestEntryKind:
+    """Tests for fs.entry_kind() classification."""
+
+    def test_regular_file(self, tmp_path):
+        path = os.path.join(str(tmp_path), "f")
+        with open(path, "w") as f:
+            f.write("x")
+        with os.scandir(str(tmp_path)) as it:
+            entry = next(iter(it))
+        assert fs.entry_kind(entry) == "file"
+
+    def test_directory(self, tmp_path):
+        path = os.path.join(str(tmp_path), "d")
+        os.mkdir(path)
+        with os.scandir(str(tmp_path)) as it:
+            entry = next(iter(it))
+        assert fs.entry_kind(entry) == "dir"
+
+    def test_symlink(self, tmp_path):
+        target = os.path.join(str(tmp_path), "target")
+        link = os.path.join(str(tmp_path), "link")
+        with open(target, "w") as f:
+            f.write("x")
+        os.symlink(target, link)
+        with os.scandir(str(tmp_path)) as it:
+            entries = {e.name: e for e in it}
+        assert fs.entry_kind(entries["link"]) == "symlink"
+
+    @pytest.mark.skipif(not hasattr(os, "mkfifo"),
+                        reason="os.mkfifo not available on this platform")
+    def test_fifo(self, tmp_path):
+        path = os.path.join(str(tmp_path), "fifo")
+        os.mkfifo(path)
+        with os.scandir(str(tmp_path)) as it:
+            entry = next(iter(it))
+        assert fs.entry_kind(entry) == "fifo"
+
+    @pytest.mark.skipif(not os.path.exists("/dev/null"),
+                        reason="no /dev/null on this platform")
+    def test_char_device(self):
+        entry = fs.PseudoDirEntry("/dev/null")
+        assert fs.entry_kind(entry) == "char_device"
+
+    @pytest.mark.skipif(not hasattr(socket, "AF_UNIX"),
+                        reason="AF_UNIX not available on this platform")
+    def test_socket(self):
+        import tempfile
+        # AF_UNIX paths are capped at ~104-108 bytes; pytest's tmp_path
+        # nests too deep on macOS, so use a short-path tempdir directly.
+        with tempfile.TemporaryDirectory() as d:
+            sock_path = os.path.join(d, "s")
+            with socket.socket(socket.AF_UNIX) as sock:
+                sock.bind(sock_path)
+                with os.scandir(d) as it:
+                    entry = next(iter(it))
+                assert fs.entry_kind(entry) == "socket"
+
+
+class TestRsyncExcludesUnsupportedEntries:
+    """
+    Regression tests: rsync() must classify FIFOs/sockets/devices via
+    lstat() alone (never opening them) and exclude them with a visible
+    EXCLUDE action, instead of either blocking forever (a FIFO with no
+    writer) or treating them as a fatal Actions.ERROR that aborts the
+    whole backup.
+    """
+
+    @pytest.mark.skipif(not hasattr(os, "mkfifo"),
+                        reason="os.mkfifo not available on this platform")
+    def test_fifo_without_writer_does_not_block(self, common_fs_dirs):
+        src_dir, dst_dir = common_fs_dirs
+        fifo_path = os.path.join(str(src_dir), "a_fifo")
+        os.mkfifo(fifo_path)
+
+        # No writer is ever opened on the other end. If rsync() opened
+        # this for reading, this test would hang instead of failing.
+        results = list(fs.rsync(str(src_dir), str(dst_dir)))
+
+        assert ("a_fifo", fs.Actions.EXCLUDE, "fifo") in results
+        assert not os.path.lexists(os.path.join(str(dst_dir), "a_fifo"))
+
+    @pytest.mark.skipif(not hasattr(socket, "AF_UNIX"),
+                        reason="AF_UNIX not available on this platform")
+    def test_socket_is_excluded_not_fatal(self):
+        import tempfile
+        with tempfile.TemporaryDirectory() as d:
+            src_dir = os.path.join(d, "src")
+            dst_dir = os.path.join(d, "dst")
+            os.mkdir(src_dir)
+            os.mkdir(dst_dir)
+            sock_path = os.path.join(src_dir, "a_socket")
+            with socket.socket(socket.AF_UNIX) as sock:
+                sock.bind(sock_path)
+                results = list(fs.rsync(src_dir, dst_dir))
+
+            assert ("a_socket", fs.Actions.EXCLUDE, "socket") in results
+            assert fs.Actions.ERROR not in [r[1] for r in results]
+            assert not os.path.lexists(os.path.join(dst_dir, "a_socket"))