Pathlib object handling for Universe, SingleFrameReaderBase and Toplogy parsers (Issue #3937)

package/CHANGELOG

@@ -34,6 +34,8 @@ Enhancements
    (Issue #4679, PR #4745)
 
 Changes
+ * Addition of Pathlib object handling for Universe, SingleFrameReaderBase
+   and Toplogy parsers (Issue #3937, PR#4535)
  * The msd.py inside analysis is changed, and ProgressBar is implemented inside
    _conclude_simple and _conclude_fft functions instead of tqdm (Issue #5144, PR #5153)
 

package/MDAnalysis/coordinates/base.py

@@ -128,6 +128,7 @@
 import abc
 import numpy as np
 import numbers
+import os
 import warnings
 from typing import Any, Union, Optional, List, Dict
 
@@ -1482,8 +1483,11 @@ def __init__(self, filename, convert_units=True, **kwargs):
 
         if isinstance(filename, NamedStream):
             self.filename = filename
+        elif isinstance(filename, (str, bytes, os.PathLike)):
+            self.filename = os.fspath(filename)  # handles Path → str
         else:
-            self.filename = str(filename)
+            # Leave non-path objects (decoders, structures, etc.) alone
+            self.filename = filename
         self.convert_units = convert_units
 
         ts_kwargs = {}
@@ -1667,7 +1671,18 @@ class SingleFrameReaderBase(ProtoReader):
     def __init__(self, filename, convert_units=True, n_atoms=None, **kwargs):
         super(SingleFrameReaderBase, self).__init__()
 
-        self.filename = filename
+        if isinstance(filename, NamedStream):
+            self.filename = filename
+
+        # Real filename types we’re okay with normalizing
+        elif isinstance(filename, (str, bytes, os.PathLike)):
+            # os.fspath handles Path objects correctly
+            self.filename = os.fspath(filename)
+
+        # Everything else (e.g. mmtf.MMTFDecoder, ParmEd structures, etc.)
+        # must be kept as-is so parsers can detect them by type.
+        else:
+            self.filename = filename
         self.convert_units = convert_units
 
         self.n_frames = 1
@@ -2310,4 +2325,4 @@ def step(self):
             Step size for iteration. Always a positive integer greater than 0.
 
         """
-        return self._step
+        return self._step

package/MDAnalysis/core/universe.py

@@ -61,6 +61,7 @@
 import contextlib
 import collections
 import inspect
+import pathlib
 
 import MDAnalysis
 import sys
@@ -219,6 +220,9 @@ def _check_file_like(topology):
         else:
             _name = None
         return NamedStream(topology, _name)
+
+    elif isinstance(topology, pathlib.Path):
+        return str(topology)
     return topology
 
 

package/MDAnalysis/topology/base.py

@@ -41,6 +41,7 @@
 import itertools
 import numpy as np
 import warnings
+import pathlib
 
 from .. import _PARSERS, _PARSER_HINTS
 from ..coordinates.base import IOBase
@@ -120,7 +121,13 @@ class TopologyReaderBase(IOBase, metaclass=_Topologymeta):
     """
 
     def __init__(self, filename):
-        self.filename = filename
+        if isinstance(filename, util.NamedStream):  # Cover NamedStream case
+            self.filename = filename
+
+        elif isinstance(filename, pathlib.Path):  # Cover Pathlib case
+            self.filename = str(filename)
+        else:
+            self.filename = filename  # Cover remainder of cases
 
     def parse(self, **kwargs):  # pragma: no cover
         raise NotImplementedError("Override this in each subclass")

testsuite/MDAnalysisTests/coordinates/base.py

@@ -25,6 +25,7 @@
 
 import numpy as np
 import pytest
+from pathlib import Path
 from unittest import TestCase
 from numpy.testing import (
     assert_equal,
@@ -35,6 +36,7 @@
 
 import MDAnalysis as mda
 from MDAnalysis.coordinates.timestep import Timestep
+from MDAnalysis.coordinates.memory import MemoryReader
 from MDAnalysis.transformations import translate
 
 
@@ -171,6 +173,12 @@ def test_pickle_singleframe_reader(self):
             "Modification of ts not preserved after serialization",
         )
 
+    def test_pathlib_input_single(self):
+        path = Path(self.filename)
+        u_str = mda.Universe(self.filename)
+        u_path = mda.Universe(path)
+        assert u_str.atoms.n_atoms == u_path.atoms.n_atoms
+
 
 class BaseReference(object):
     def __init__(self):
@@ -667,6 +675,18 @@ def test_timeseries_atomgroup_asel_mutex(self, reader):
                 atomgroup=atoms, asel=atoms, order="fac"
             )
 
+    def test_pathlib_input(self, reader):
+        if isinstance(reader, MemoryReader):
+            if isinstance(reader, MemoryReader):
+                skip_reason = "MemoryReader"
+            pytest.skip(
+                f"Skipping test for Pathlib input with reason: {skip_reason}"
+            )
+        path = Path(reader.filename)
+        u_str = mda.Universe(reader.filename)
+        u_path = mda.Universe(path)
+        assert u_str.atoms.n_atoms == u_path.atoms.n_atoms
+
 
 class MultiframeReaderTest(BaseReaderTest):
     def test_last_frame(self, ref, reader):

testsuite/MDAnalysisTests/coordinates/test_gro.py

@@ -48,7 +48,6 @@
     assert_equal,
 )
 import pytest
-from io import StringIO
 
 
 class TestGROReaderOld(RefAdK):

testsuite/MDAnalysisTests/coordinates/test_imd.py

@@ -6,6 +6,8 @@
 import sys
 from types import ModuleType
 from weakref import ref
+from pathlib import Path
+from MDAnalysis.topology.MinimalParser import MinimalParser
 
 import pytest
 import numpy as np
@@ -323,6 +325,25 @@ def test_volume(self, ref, reader):
         # to limited floating point precision.
         assert_allclose(vol, ref.volume, rtol=0, atol=1.5e0)
 
+    def test_pathlib_input(self, ref, reader):
+        """
+        IMDReader cannot deduce n_atoms from the IMD URL, so we bypass
+        Universe(...) and directly test MinimalParser with both
+        str and pathlib.Path inputs, passing n_atoms explicitly.
+        """
+        imd_uri = reader.filename
+        path = Path(imd_uri)
+
+        with MinimalParser(imd_uri) as p:
+            top_str = p.parse(n_atoms=ref.n_atoms)
+
+        with MinimalParser(path) as p:
+            top_path = p.parse(n_atoms=ref.n_atoms)
+
+        assert top_str.n_atoms == ref.n_atoms
+        assert top_path.n_atoms == ref.n_atoms
+        assert top_str.n_atoms == top_path.n_atoms
+
     def test_reload_auxiliaries_from_description(self, ref, reader):
         pytest.skip("Cannot create two IMDReaders on the same stream")
 

testsuite/MDAnalysisTests/topology/base.py

@@ -21,6 +21,7 @@
 # J. Comput. Chem. 32 (2011), 2319--2327, doi:10.1002/jcc.21787
 #
 import pytest
+import pathlib
 
 import MDAnalysis as mda
 from MDAnalysis.core.topology import Topology
@@ -123,3 +124,21 @@ def test_guessed_attributes(self, filename):
         for attr in self.guessed_attrs:
             assert hasattr(u.atoms, attr)
             assert attr in u_guessed_attrs
+
+    def test_pathlib_input(self, filename):
+        """Check that pathlib.Path objects are accepted by the parser."""
+        if not isinstance(filename, (str, pathlib.Path)):
+            pytest.skip(
+                f"Pathlib input test only applies to string/path-like filenames, "
+                f"got {type(filename).__name__}"
+            )  # Cover OpenMM Parser case
+        path = pathlib.Path(filename)
+
+        with self.parser(filename) as p:
+            top_str = p.parse()
+        with self.parser(path) as p:
+            top_path = p.parse()
+
+        assert top_str.n_atoms == top_path.n_atoms
+        assert top_str.n_residues == top_path.n_residues
+        assert top_str.n_segments == top_path.n_segments

testsuite/MDAnalysisTests/topology/test_itp.py

@@ -493,6 +493,18 @@ def test_sequential(self, universe):
         )
         assert_equal(universe.atoms.chargegroups[-1], 63)
 
+    def test_pathlib_input(self, filename):
+        path = Path(filename)
+
+        with self.parser(filename) as p:
+            top_str = p.parse(include_dir=GMX_DIR)
+        with self.parser(path) as p:
+            top_path = p.parse(include_dir=GMX_DIR)
+
+        assert top_str.n_atoms == top_path.n_atoms
+        assert top_str.n_residues == top_path.n_residues
+        assert top_str.n_segments == top_path.n_segments
+
 
 class TestErrors: