iterables

* allow any kind of iterables

* sqlalchemy not needed anymore

* test on python3.8

Author: eloyfelix

.travis.yml

@@ -6,10 +6,14 @@ matrix:
     env: CONDA=3.6
   - os: linux
     env: CONDA=3.7
+  - os: linux
+    env: CONDA=3.8
   - os: osx
     env: CONDA=3.6
   - os: osx
     env: CONDA=3.7
+  - os: osx
+    env: CONDA=3.8
 
 before_install:
 - |

FPSim2/io/FPSim2_io.py

@@ -1,10 +1,10 @@
 import rdkit
 from rdkit import Chem
-from sqlalchemy.engine.result import ResultProxy
 from FPSim2.FPSim2lib import py_popcount
 from rdkit.Chem import rdMolDescriptors
 from rdkit.Avalon import pyAvalonTools
 from collections import namedtuple
+from collections.abc import Iterable
 import tables as tb
 import numpy as np
 import textwrap
@@ -336,21 +336,21 @@ def get_mol_suplier(io_source):
     """Returns a mol supplier depending on the object type and file extension.
 
     Args:
-        io_source: source of molecules, smi or sdf filenames, 
-                   SQLA ResultProxy or python list.
+        io_source: source of molecules; smi or sdf filenames, 
+                   or any iterable object.
     Returns:
         molecule supplier generator.
     """
     supplier = None
-    if isinstance(io_source, list) or isinstance(io_source, ResultProxy):
-        supplier = it_supplier
-    else:
+    if isinstance(io_source, str):
         input_type = io_source.split(".")[-1]
         if input_type == "smi":
             supplier = smi_mol_supplier
         elif input_type == "sdf":
             supplier = sdf_mol_supplier
-    if not supplier:
+    elif isinstance(io_source, Iterable):
+        supplier = it_supplier
+    else:
         raise Exception("No valid input molecules input.")
     return supplier
 

README.md

@@ -24,16 +24,7 @@ From source:
  - clone the repo
  - `pip install ./FPSim2`
 
-From a conda environment. Builds available for:
-- linux:
-    - Python 3.6
-    - Python 3.7
-- mac:
-    - Python 3.6
-    - Python 3.7
-- windows:
-    - Python 3.6
-    - Python 3.7
+From a conda environment:
 
 ```bash
 conda install -c efelix fpsim2
@@ -55,7 +46,7 @@ create_db_file('chembl.sdf', 'chembl.h5', 'Morgan', {'radius': 2, 'nBits': 2048}
 # from Python list
 create_db_file([['CC', 1], ['CCC', 2], ['CCCC', 3]], 'test/10mols.h5', 'Morgan', {'radius': 2, 'nBits': 2048})
 
-# from sqlalchemy ResulProxy
+# or any other iterable like sqlalchemy ResultProxy 
 from sqlalchemy.orm import Session
 from sqlalchemy import create_engine
 
@@ -97,7 +88,7 @@ In case RDKit is not able to load a molecule, the id assigned to the molecule wi
 ```python
 from FPSim2 import FPSim2Engine
 
-fp_filename = 'chembl_25.h5'
+fp_filename = 'chembl_27.h5'
 query = 'CC(=O)Oc1ccccc1C(=O)O'
 
 fpe = FPSim2Engine(fp_filename)
@@ -114,7 +105,7 @@ If you're searching against a huge dataset or you have small RAM, you can still
 ```python
 from FPSim2 import FPSim2Engine
 
-fp_filename = 'chembl_25.h5'
+fp_filename = 'chembl_27.h5'
 query = 'CC(=O)Oc1ccccc1C(=O)O'
 
 fpe = FPSim2Engine(fp_filename, in_memory_fps=False)

conda.recipe/meta.yaml

@@ -22,7 +22,6 @@ requirements:
     - python
     - numpy>=1.14
     - pytables>=3.4.4
-    - sqlalchemy>=1.2
     - rdkit>=2018.09
     - vs2015_runtime  # [win]
 

setup.py

@@ -3,7 +3,7 @@
 import platform
 import sys
 
-__version__ = "0.1.5"
+__version__ = "0.1.6"
 
 
 class get_pybind_include(object):
@@ -121,6 +121,7 @@ def build_extensions(self):
         "License :: OSI Approved :: MIT License",
         "Programming Language :: Python :: 3.6",
         "Programming Language :: Python :: 3.7",
+        "Programming Language :: Python :: 3.8",
         "Topic :: Scientific/Engineering :: Chemistry",
     ],
 )

test/test_fpsim.py

@@ -12,8 +12,6 @@
     sort_db_file,
 )
 from rdkit import Chem, DataStructs
-from sqlalchemy import create_engine
-from sqlalchemy.orm import Session
 import unittest
 import os
 
@@ -100,22 +98,6 @@ def test_d_create_db_file_list(self):
             assert config[1]["nBits"] == fp_params["nBits"]
             assert fp_file.root.fps.shape[0] == 3
 
-    def test_e_create_db_file_sqla(self):
-        in_file = os.path.join(TESTS_DIR, 'data/test.db')
-        out_file = os.path.join(TESTS_DIR, 'data/10mols_sqla.h5')
-        engine = create_engine("sqlite:///{}".format(in_file))
-        s = Session(engine)
-        sql_query = "select mol_string, mol_id from structure"
-        resprox = s.execute(sql_query)
-
-        create_db_file(resprox, out_file, fp_type, fp_params)
-        with tb.open_file(out_file, mode="r") as fp_file:
-            config = fp_file.root.config
-            assert config[0] == fp_type
-            assert config[1]["radius"] == fp_params["radius"]
-            assert config[1]["nBits"] == fp_params["nBits"]
-            assert fp_file.root.fps.shape[0] == 10
-
     def test_f_load_fps(self):
         in_file = os.path.join(TESTS_DIR, 'data/10mols.h5')
         fps = load_fps(in_file)