Merge pull request #84 from IAlibay/streams

Allows for file-like objects to be passed to read_molecule_file

propka/input.py

@@ -20,10 +20,10 @@ def open_file_for_reading(input_file):
 
     Args:
         input_file: path to file or file-like object. If file-like object,
-        then will attempt fseek(0).
+        then will attempt seek(0).
     """
     try:
-        input_file.fseek(0)
+        input_file.seek(0)
         return input_file
     except AttributeError:
         pass
@@ -35,27 +35,62 @@ def open_file_for_reading(input_file):
     return file_
 
 
-def read_molecule_file(input_file, mol_container):
+def read_molecule_file(filename: str, mol_container, stream=None):
-    """Read input file (PDB or PROPKA) for a molecular container
+    """Read input file or stream (PDB or PROPKA) for a molecular container
 
-    Args
+    Args:
-        input_file:  input file to read
+        filename(str):  name of input file. If not using a filestream via the
-        mol_container:  MolecularContainer object
+            ``stream`` argument, should be a path to the file to be read.
-    Returns
+        mol_container:  MolecularContainer object.
+        stream: optional filestream handle. If ``None``, then open
+            ``filename`` as a local file for reading.
+
+    Returns:
         updated MolecularContainer object
-    Raises
+
+    Raises:
         ValuError if invalid input given
+
+    Examples:
+        There are two main cases for using ``read_molecule_file``. The first
+        (and most common) is to pass the input file (``filename``) as a
+        string which gives the path of the molecule file to be read (here we
+        also pass a ``MoleculeContainer`` object named ``mol_container``).
+
+        >>> read_molecule_file('test.pdb', mol_container)
+        <propka.molecular_container.MolecularContainer at 0x7f6e0c8f2310>
+
+        The other use case is when passing a file-like object, e.g. a
+        ``StringIO`` class, instance. This is done by passing the object via
+        the ``stream`` argument. Since file-like objects do not usually have
+        an associated file name, an appropirate file name should be passed to
+        the ``filename`` argument. In this case, ``filename`` is not opened for
+        reading, but instead is used to help recognise the file type (based on
+        the extension being either `.pdb` or `.propka_input`) and also uses
+        that given ``filename`` to assign a name to the input
+        MolecularContainer object.
+
+        >>> read_molecule_file('test.pdb', mol_container,
+                               stream=string_io_object)
+        <propka.molecular_container.MolecularContainer at 0x7f6e0c8f2310>
+
     """
-    input_path = Path(input_file)
+    input_path = Path(filename)
     mol_container.name = input_path.stem
     input_file_extension = input_path.suffix
+
+    if stream is not None:
+        input_file = stream
+    else:
+        input_file = filename
+
     if input_file_extension.lower() == '.pdb':
         # input is a pdb file. read in atoms and top up containers to make
         # sure that all atoms are present in all conformations
         conformations, conformation_names = read_pdb(
-            input_path, mol_container.version.parameters, mol_container)
+            input_file, mol_container.version.parameters, mol_container)
         if len(conformations) == 0:
-            str_ = ('Error: The pdb file does not seems to contain any '
+            str_ = ('Error: The pdb file does not seem to contain any '
                     'molecular conformations')
             raise ValueError(str_)
         mol_container.conformations = conformations

tests/__init__.py

@@ -0,0 +1 @@
+

tests/test_streamio.py

@@ -0,0 +1,120 @@
+"""Tests for PROPKA stream io"""
+import logging
+from pathlib import Path
+from io import StringIO
+import pytest
+from propka.parameters import Parameters
+from propka.molecular_container import MolecularContainer
+from propka.input import read_parameter_file, read_molecule_file
+from propka.lib import loadOptions
+
+from .test_basic_regression import get_test_dirs, compare_output
+
+
+_LOGGER = logging.getLogger(__name__)
+
+
+def get_paths(pdb):
+    """Helper function to get the path to the input and reference files"""
+    path_dict = get_test_dirs()
+    ref_path = path_dict["results"] / ("{0:s}.dat".format(pdb))
+    pdb_path = path_dict["pdbs"] / ("{0:s}.pdb".format(pdb))
+
+    return ref_path.resolve(), pdb_path.resolve()
+
+
+def run_propka_stream(options, input_file, filename):
+    """Run PROPKA software.
+
+    Args:
+        options:  list of PROPKA options
+        input_file:  file-like PDB object
+        filename: filename for the file-like PDB object
+    """
+    options += [filename]
+    args = loadOptions(options)
+    parameters = read_parameter_file(args.parameters, Parameters())
+    molecule = MolecularContainer(parameters, args)
+    molecule = read_molecule_file(filename, molecule, stream=input_file)
+    molecule.calculate_pka()
+    molecule.write_pka()
+    if args.generate_propka_input:
+        molecule.write_propka()
+
+
+@pytest.mark.parametrize("pdb, options", [
+    pytest.param("1FTJ-Chain-A", [], id="1FTJ-Chain-A: no options"),
+    pytest.param('3SGB-subset', [
+        "--titrate_only",
+        "E:17,E:18,E:19,E:29,E:44,E:45,E:46,E:118,E:119,E:120,E:139"],
+                 id="3SGB: --titrate_only"),
+    pytest.param('1HPX-warn', ['--quiet'], id="1HPX-warn: --quiet"),
+])
+def test_textio_filestream(tmpdir, pdb, options):
+    """Basic regression test using TextIO streams for the input PDB file"""
+    # Get the relevant paths
+    ref_path, pdb_path = get_paths(pdb)
+    filename = f"{pdb}.pdb"
+
+    filestream = open(pdb_path, 'r')
+
+    with tmpdir.as_cwd():
+        run_propka_stream(options, filestream, filename)
+        compare_output(pdb, Path.cwd(), ref_path)
+
+    filestream.close()
+
+
+@pytest.mark.parametrize("pdb, options", [
+    pytest.param("1FTJ-Chain-A", [], id="1FTJ-Chain-A: no options"),
+    pytest.param('3SGB-subset', [
+        "--titrate_only",
+        "E:17,E:18,E:19,E:29,E:44,E:45,E:46,E:118,E:119,E:120,E:139"],
+                 id="3SGB: --titrate_only"),
+    pytest.param('1HPX-warn', ['--quiet'], id="1HPX-warn: --quiet"),
+])
+def test_stringio_filestream(tmpdir, pdb, options):
+    """Basic regression test using StringIO streams for the input PDB file"""
+    # Get the relevant paths
+    ref_path, pdb_path = get_paths(pdb)
+    filename = f"{pdb}.pdb"
+
+    with open(pdb_path, 'r') as writer:
+        filestream = StringIO(writer.read())
+
+    with tmpdir.as_cwd():
+        run_propka_stream(options, filestream, filename)
+        compare_output(pdb, Path.cwd(), ref_path)
+
+    filestream.close()
+
+
+def test_valuerror_nofiletype():
+    """Tests for raised ValueError when an unknown filename is passed to
+    read_molecule_file"""
+    pdb = "1FTJ-Chain-A"
+    options = []
+
+    ref_path, pdb_path = get_paths(pdb)
+
+    with open(pdb_path, 'r') as writer:
+        filestream = StringIO(writer.read())
+
+    errmsg = "Unknown input file type"
+    with pytest.raises(ValueError, match=errmsg):
+        run_propka_stream(options, filestream, filename="test.dat")
+
+
+def test_valuerror_notpdb():
+    """Tests for raised ValueError when a stream object that isn't a PDB
+    is passed to read_molecule_file"""
+    pdb = "1FTJ-Chain-A"
+    options = []
+
+    ref_path, pdb_path = get_paths(pdb)
+
+    filestream = StringIO()
+
+    errmsg = "The pdb file does not seem to contain any "
+    with pytest.raises(ValueError, match=errmsg):
+        run_propka_stream(options, filestream, filename="test.pdb")