Update TarArchiveOperations and equip with tests

.appveyor.yml

@@ -202,6 +202,9 @@ build: off
 # init cannot use any components from the repo, because it runs prior to
 # cloning it
 init:
+  # enable external SSH access to CI worker
+  # needs APPVEYOR_SSH_KEY defined in project settings (or environment)
+  - sh: curl -sflL 'https://raw.githubusercontent.com/appveyor/ci/master/scripts/enable-ssh.sh' | bash -e -
   # remove windows 260-char limit on path names
   - cmd: powershell Set-Itemproperty -path "HKLM:\SYSTEM\CurrentControlSet\Control\FileSystem" -Name LongPathsEnabled -value 1
   # enable developer mode on windows
@@ -227,9 +230,6 @@ init:
   - sh: export TMPDIR=~/DLTMP
 
 install:
-  # enable external SSH access to CI worker on all other systems
-  # needs APPVEYOR_SSH_KEY defined in project settings (or environment)
-  - sh: tools/appveyor/enable-ssh-login
   # place a debug setup helper at a convenient location
   - cmd: copy tools\appveyor\env_setup.bat C:\\datalad_debug.bat
   # Missing system software

datalad_next/archive_operations/__init__.py

@@ -1,6 +1,6 @@
 """Handler for operations on various archive types
 
-All handlers implements the API defined by :class:`ArchiveOperations`.
+All handlers implement the API defined by :class:`ArchiveOperations`.
 
 Available handlers:
 
@@ -71,6 +71,10 @@ def __init__(self, location: Any, *, cfg: ConfigManager | None = None):
     def __str__(self) -> str:
         return f'{self.__class__.__name__}({self._location})'
 
+    def __repr__(self) -> str:
+        return \
+            f'{self.__class__.__name__}({self._location}, cfg={self._cfg!r})'
+
     @property
     def cfg(self) -> ConfigManager:
         """ConfigManager given to the constructor, or the session default"""

datalad_next/archive_operations/tarfile.py

@@ -6,13 +6,17 @@
 import logging
 import tarfile
 from contextlib import contextmanager
-from pathlib import Path
+from pathlib import (
+    Path,
+    PurePath,
+)
 from typing import (
     Any,
     Generator,
     IO,
 )
 
+from datalad_next.config import ConfigManager
 # TODO we might just want to do it in reverse:
 # move the code of `iter_tar` in here and have it call
 # `TarArchiveOperations(path).__iter__()` instead.
@@ -27,13 +31,15 @@
 )
 
 from . import ArchiveOperations
-from datalad_next.config import ConfigManager
 
 lgr = logging.getLogger('datalad.ext.next.archive_operations.tarfile')
 
 
 class TarArchiveOperations(ArchiveOperations):
-    """
+    """Handler for a TAR archive on a local file system
+
+    Any methods that take an archive item/member name as an argument
+    accept a POSIX path string, or any `PurePath` instance.
     """
     def __init__(self, location: Path, *, cfg: ConfigManager | None = None):
         """
@@ -55,24 +61,36 @@ def __init__(self, location: Path, *, cfg: ConfigManager | None = None):
 
     @property
     def tarfile(self) -> tarfile.TarFile:
+        """Returns `TarFile` instance, after creating it on-demand
+
+        The instance is cached, and needs to be released by calling
+        ``.close()`` if called outside a context manager.
+        """
         if self._tarfile is None:
             self._tarfile = tarfile.open(self._tarfile_path, 'r')
         return self._tarfile
 
     def close(self) -> None:
+        """Closes any opened TAR file handler"""
         if self._tarfile:
             self._tarfile.close()
             self._tarfile = None
 
     @contextmanager
-    def open(self, item: Any) -> IO:
-        """
+    def open(self, item: str | PurePath) -> IO:
+        """Get a file-like for a TAR archive item
+
+        Parameters
+        ----------
+        item: str | PurePath
+          The identifier must be a POSIX path string, or a `PurePath` instance.
         """
-        yield self.tarfile.extractfile(str(item))
+        with self.tarfile.extractfile(_anyid2membername(item)) as fp:
+            yield fp
 
-    def __contains__(self, item: Any) -> bool:
+    def __contains__(self, item: str | PurePath) -> bool:
         try:
-            self.tarfile.getmember(item)
+            self.tarfile.getmember(_anyid2membername(item))
             return True
         except KeyError:
             return False
@@ -81,3 +99,10 @@ def __iter__(self) -> Generator[TarfileItem, None, None]:
         # if fp=True is needed, either `iter_tar()` can be used
         # directly, or `TarArchiveOperations.open`
         yield from iter_tar(self._tarfile_path, fp=False)
+
+
+def _anyid2membername(item_id: str | PurePath) -> str:
+    if isinstance(item_id, PurePath):
+        return item_id.as_posix()
+    else:
+        return item_id

datalad_next/archive_operations/tests/test_tarfile.py

@@ -0,0 +1,87 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from pathlib import (
+    Path,
+    PurePath,
+    PurePosixPath,
+)
+from typing import Generator
+
+import pytest
+
+from datalad_next.iter_collections.utils import FileSystemItemType
+
+from ..tarfile import TarArchiveOperations
+
+
+@dataclass
+class TestArchive:
+    path: Path
+    item_count: int
+    content: bytes
+    target_hash: dict[str, str]
+
+
+@pytest.fixture(scope='session')
+def structured_sample_tar_xz(
+    sample_tar_xz
+) -> Generator[TestArchive, None, None]:
+    yield TestArchive(
+        path=sample_tar_xz,
+        item_count=6,
+        content=b'123\n',
+        target_hash={
+            'SHA1': 'b5dfcec4d1b6166067226fae102f7fbcf6bd1bd4',
+            'md5': 'd700214df5487801e8ee23d31e60382a',
+        }
+    )
+
+
+def test_tararchive_basics(structured_sample_tar_xz: TestArchive):
+    spec = structured_sample_tar_xz
+    # this is intentionally a hard-coded POSIX relpath
+    member_name = 'test-archive/onetwothree.txt'
+    with TarArchiveOperations(spec.path) as archive_ops:
+        with archive_ops.open(member_name) as member:
+            assert member.read() == spec.content
+        with archive_ops.open(PurePosixPath(member_name)) as member:
+            assert member.read() == spec.content
+        with archive_ops.open(PurePath(member_name)) as member:
+            assert member.read() == spec.content
+
+
+def test_tararchive_contain(structured_sample_tar_xz: TestArchive):
+    # this is intentionally a hard-coded POSIX relpath
+    member_name = 'test-archive/onetwothree.txt'
+    archive_ops = TarArchiveOperations(structured_sample_tar_xz.path)
+    # POSIX path str
+    assert member_name in archive_ops
+    # POSIX path as obj
+    assert PurePosixPath(member_name) in archive_ops
+    # platform path
+    assert PurePath(PurePosixPath(member_name)) in archive_ops
+    assert 'bogus' not in archive_ops
+
+
+def test_tararchive_iterator(structured_sample_tar_xz: TestArchive):
+    spec = structured_sample_tar_xz
+    with TarArchiveOperations(spec.path) as archive_ops:
+        items = list(archive_ops)
+        assert len(items) == spec.item_count
+        for item in items:
+            assert item.name in archive_ops
+
+
+def test_open(structured_sample_tar_xz: TestArchive):
+    spec = structured_sample_tar_xz
+    file_pointer = set()
+    with TarArchiveOperations(spec.path) as tf:
+        for item in tf:
+            if item.type == FileSystemItemType.file:
+                with tf.open(str(PurePosixPath(item.name))) as fp:
+                    file_pointer.add(fp)
+                    assert fp.read(len(spec.content)) == spec.content
+        # check the fp before we close the archive handler
+        for fp in file_pointer:
+            assert fp.closed is True