Source Code for Module rdkit.Chem.ChemUtils.BulkTester

 1  # $Id: BulkTester.py 1528 2010-09-26 17:04:37Z glandrum $ 
 2  # 
 3  #  Copyright (C) 2005-2006 Rational Discovery LLC 
 4  # 
 5  #   @@ All Rights Reserved @@ 
 6  #  This file is part of the RDKit. 
 7  #  The contents are covered by the terms of the BSD license 
 8  #  which is included in the file license.txt, found at the root 
 9  #  of the RDKit source tree. 
10  # 
11  from rdkit import Chem 
12  import sys 
13  from rdkit.Chem import Randomize 
14   
15 -def TestMolecule(mol): 
16    try: 
17      Chem.SanitizeMol(mol) 
18      mol = Chem.RemoveHs(mol) 
19    except ValueError,msg: 
20      return -1 
21    except: 
22      import traceback 
23      traceback.print_exc() 
24      return -2 
25    if mol.GetNumAtoms(): 
26      try: 
27        Randomize.CheckCanonicalization(mol,10) 
28      except: 
29        import traceback 
30        traceback.print_exc() 
31        return -3 
32    return 0 
33       
34       
35 -def TestSupplier(suppl,stopAfter=-1,reportInterval=100,reportTo=sys.stderr, 
36                   nameProp='_Name'): 
37    nDone = 0 
38    nFailed = 0 
39    while 1: 
40      try: 
41        mol = suppl.next() 
42      except StopIteration: 
43        break 
44      except: 
45        import traceback 
46        traceback.print_exc() 
47        nFailed += 1 
48        reportTo.flush() 
49        print >>reportTo,'Failure at mol %d'%nDone 
50      else: 
51        if mol: 
52          ok = TestMolecule(mol) 
53        else: 
54          ok = -3 
55        if ok<0: 
56          nFailed += 1 
57          reportTo.flush() 
58          if ok==-3: 
59            print >>reportTo,'Canonicalization', 
60          print >>reportTo,'Failure at mol %d'%nDone, 
61          if mol: 
62            print >>reportTo,mol.GetProp(nameProp), 
63          print >>reportTo,'' 
64   
65             
66      nDone += 1 
67      if nDone==stopAfter: 
68        break 
69      if not nDone%reportInterval: 
70        print 'Done %d molecules, %d failures'%(nDone,nFailed) 
71    return nDone,nFailed 
72  if __name__=='__main__': 
73    suppl = Chem.SDMolSupplier(sys.argv[1],False) 
74    if len(sys.argv)>2: 
75      nameProp = sys.argv[2] 
76    else: 
77      nameProp = '_Name' 
78       
79    nDone,nFailed = TestSupplier(suppl,nameProp=nameProp) 
80    print '%d failures in %d mols'%(nFailed,nDone) 
81