Support pathlib paths in HDMFIO/HDF5IO (#450)

* Fix #439
* Ref NeurodataWithoutBorders/pynwb#1303
* Support pathlib.Path as source parameter in HDMFIO.__init__
* Support pathlib.Path as path parameter in HDF5IO.__init__
* Suport pathlib.Path as path parameter in HDF5IO.load_namespaces
* Add new tests using pathlib paths

CHANGELOG.md

@@ -11,6 +11,7 @@
 - Add capability to add a row to a column after IO. @bendichter (#426)
 - Add method `hdmf.utils.get_docval_macro` to get a tuple of the current values for a docval_macro, e.g., 'array_data'  
   and 'scalar_data'. @rly (#456)
+- Support `pathlib.Path` paths in `HDMFIO.__init__`, `HDF5IO.__init__`, and `HDF5IO.load_namespaces`. @dsleiter (#439)
 
 ### Internal improvements
 - Refactor `HDF5IO.write_dataset` to be more readable. @rly (#428)

src/hdmf/backends/hdf5/h5tools.py

@@ -1,6 +1,7 @@
 from collections import deque
 import numpy as np
 import os.path
+from pathlib import Path
 from functools import partial
 from h5py import File, Group, Dataset, special_dtype, SoftLink, ExternalLink, Reference, RegionReference, check_dtype
 import logging
@@ -29,7 +30,7 @@
 
 class HDF5IO(HDMFIO):
 
-    @docval({'name': 'path', 'type': str, 'doc': 'the path to the HDF5 file'},
+    @docval({'name': 'path', 'type': (str, Path), 'doc': 'the path to the HDF5 file'},
             {'name': 'manager', 'type': (TypeMap, BuildManager),
              'doc': 'the BuildManager or a TypeMap to construct a BuildManager to use for I/O', 'default': None},
             {'name': 'mode', 'type': str,
@@ -45,6 +46,9 @@ def __init__(self, **kwargs):
         self.logger = logging.getLogger('%s.%s' % (self.__class__.__module__, self.__class__.__qualname__))
         path, manager, mode, comm, file_obj = popargs('path', 'manager', 'mode', 'comm', 'file', kwargs)
 
+        if isinstance(path, Path):
+            path = str(path)
+
         if file_obj is not None and os.path.abspath(file_obj.filename) != os.path.abspath(path):
             msg = 'You argued %s as this object\'s path, ' % path
             msg += 'but supplied a file with filename: %s' % file_obj.filename
@@ -86,7 +90,7 @@ def _file(self):
     @classmethod
     @docval({'name': 'namespace_catalog', 'type': (NamespaceCatalog, TypeMap),
              'doc': 'the NamespaceCatalog or TypeMap to load namespaces into'},
-            {'name': 'path', 'type': str, 'doc': 'the path to the HDF5 file', 'default': None},
+            {'name': 'path', 'type': (str, Path), 'doc': 'the path to the HDF5 file', 'default': None},
             {'name': 'namespaces', 'type': list, 'doc': 'the namespaces to load', 'default': None},
             {'name': 'file', 'type': File, 'doc': 'a pre-existing h5py.File object', 'default': None},
             returns="dict with the loaded namespaces", rtype=dict)
@@ -100,6 +104,9 @@ def load_namespaces(cls, **kwargs):
         namespace_catalog, path, namespaces, file_obj = popargs('namespace_catalog', 'path', 'namespaces', 'file',
                                                                 kwargs)
 
+        if isinstance(path, Path):
+            path = str(path)
+
         if path is None and file_obj is None:
             raise ValueError("Either the 'path' or 'file' argument must be supplied to load_namespaces.")
 

src/hdmf/backends/io.py

@@ -1,4 +1,5 @@
 from abc import ABCMeta, abstractmethod
+from pathlib import Path
 
 from ..build import BuildManager, GroupBuilder
 from ..utils import docval, getargs, popargs
@@ -8,12 +9,16 @@
 class HDMFIO(metaclass=ABCMeta):
     @docval({'name': 'manager', 'type': BuildManager,
              'doc': 'the BuildManager to use for I/O', 'default': None},
-            {"name": "source", "type": str,
+            {"name": "source", "type": (str, Path),
              "doc": "the source of container being built i.e. file path", 'default': None})
     def __init__(self, **kwargs):
-        self.__manager = getargs('manager', kwargs)
+        manager, source = getargs('manager', 'source', kwargs)
+        if isinstance(source, Path):
+            source = str(source)
+
+        self.__manager = manager
         self.__built = dict()
-        self.__source = getargs('source', kwargs)
+        self.__source = source
         self.open()
 
     @property

tests/unit/test_io_hdf5_h5tools.py

@@ -4,6 +4,7 @@
 import numpy as np
 import h5py
 from io import BytesIO
+from pathlib import Path
 
 from hdmf.utils import docval, getargs
 from hdmf.data_utils import DataChunkIterator, InvalidDataIOError
@@ -850,6 +851,20 @@ def test_roundtrip_empty_group(self):
             read_foofile = io.read()
             self.assertDictEqual({}, read_foofile.buckets['bucket1'].foos)
 
+    def test_roundtrip_pathlib_path(self):
+        pathlib_path = Path(self.path)
+        foo1 = Foo('foo1', [1, 2, 3, 4, 5], "I am foo1", 17, 3.14)
+        foobucket = FooBucket('bucket1', [foo1])
+        foofile = FooFile([foobucket])
+
+        with HDF5IO(pathlib_path, manager=self.manager, mode='w') as io:
+            io.write(foofile)
+
+        with HDF5IO(pathlib_path, manager=self.manager, mode='r') as io:
+            read_foofile = io.read()
+            self.assertListEqual(foofile.buckets['bucket1'].foos['foo1'].my_data,
+                                 read_foofile.buckets['bucket1'].foos['foo1'].my_data[:].tolist())
+
 
 class TestHDF5IO(TestCase):
 
@@ -885,6 +900,11 @@ def test_set_file_mismatch(self):
         with self.assertRaisesWith(ValueError, err_msg):
             HDF5IO(self.path, manager=self.manager, mode='w', file=self.file_obj)
 
+    def test_pathlib_path(self):
+        pathlib_path = Path(self.path)
+        with HDF5IO(pathlib_path, mode='w') as io:
+            self.assertEqual(io.source, self.path)
+
 
 class TestCacheSpec(TestCase):
 
@@ -1988,6 +2008,22 @@ def test_load_namespaces_file_path_mismatched(self):
 
         file_obj.close()
 
+    def test_load_namespaces_with_pathlib_path(self):
+        """Test that loading a namespace using a valid pathlib Path is OK and returns the correct dictionary."""
+
+        # Setup all the data we need
+        foo1 = Foo('foo1', [1, 2, 3, 4, 5], "I am foo1", 17, 3.14)
+        foobucket = FooBucket('bucket1', [foo1])
+        foofile = FooFile([foobucket])
+
+        with HDF5IO(self.path, manager=self.manager, mode='w') as io:
+            io.write(foofile)
+
+        pathlib_path = Path(self.path)
+        ns_catalog = NamespaceCatalog()
+        d = HDF5IO.load_namespaces(ns_catalog, pathlib_path)
+        self.assertEqual(d, {'test_core': {}})  # test_core has no dependencies
+
 
 class TestExport(TestCase):
     """Test exporting HDF5 to HDF5 using HDF5IO.export_container_to_hdf5."""