Python 3 compatibility (issue #398) (#1192)

* Replace xrange with range or import from rdkit.six

* Remove use of string.lower

* The import of xrange in MolSurf added it to the Descriptors._descList

Contrib/Glare/glare.py

@@ -205,7 +205,7 @@ class Library:
         max_subsets = max(partitions)
 
         enumeration_indices = []
-        for i in xrange(max_subsets):
+        for i in range(max_subsets):
             combinations = []
             for size in partitions:
                 combinations.append( i % size )

Contrib/fraggle/rdkit_tversky.py

@@ -97,7 +97,7 @@ for line in sys.stdin:
   res = DataStructs.BulkTverskySimilarity(mfp, queries, 0, 1, False)
 
   #query_frag_smiles,query_smiles,query_id,retrieved_smi,retrieved_id,tversky_sim
-  for i in xrange(fragments):
+  for i in range(fragments):
     if (res[i] >= options.cutoff):
       print("%s,%s,%s,%s,%s,%s" %
             (query_info[i][2], query_info[i][0], query_info[i][1], smi, id, res[i]))

rdkit/Chem/Fingerprints/ClusterMols.py

@@ -48,8 +48,8 @@ def GetDistanceMatrix(data, metric, isSimilarity=1):
   nPts = len(data)
   res = numpy.zeros((nPts * (nPts - 1) / 2), numpy.float)
   nSoFar = 0
-  for col in xrange(1, nPts):
-    for row in xrange(col):
+  for col in range(1, nPts):
+    for row in range(col):
       fp1 = data[col][1]
       fp2 = data[row][1]
       if fp1.GetNumBits() > fp2.GetNumBits():

rdkit/Chem/MolSurf.py

@@ -72,7 +72,7 @@ def _pyLabuteHelper(mol, includeHs=1, force=0):
 
   # 0 contains the H information
   rads[0] = numTable[1][radCol]
-  for i in xrange(nAts):
+  for i in range(nAts):
     rads[i + 1] = numTable[mol.GetAtomWithIdx(i).GetAtomicNum()][radCol]
 
   # start with explicit bonds
@@ -94,14 +94,14 @@ def _pyLabuteHelper(mol, includeHs=1, force=0):
   if includeHs:
     j = 0
     Rj = rads[j]
-    for i in xrange(1, nAts + 1):
+    for i in range(1, nAts + 1):
       Ri = rads[i]
       bij = Ri + Rj
       dij = min(max(abs(Ri - Rj), bij), Ri + Rj)
       Vi[i] += Rj * Rj - (Ri - dij)**2 / dij
       Vi[j] += Ri * Ri - (Rj - dij)**2 / dij
 
-  for i in xrange(nAts + 1):
+  for i in range(nAts + 1):
     Ri = rads[i]
     Vi[i] = 4 * math.pi * Ri**2 - math.pi * Ri * Vi[i]
 

rdkit/Chem/UnitTestDescriptors.py

@@ -44,7 +44,7 @@ class TestCase(unittest.TestCase):
         except Exception:
           import traceback
           traceback.print_exc()
-          raise AssertionError('SMILES: %s' % smi)
+          raise AssertionError('SMILES: %s; Descriptor: %s' % (smi, nm))
 
   def testMolFormula(self):
     for (smiles, expected) in (("[NH4+]", "H4N+"),

rdkit/Chem/fmcs/fmcs.py

@@ -1047,7 +1047,7 @@ def find_duplicates(cangen_nodes, start, end):
   result = []
   prev_value = -1
   count = 0
-  for index in xrange(start, end):
+  for index in range(start, end):
     node = cangen_nodes[index]
     if node.value == prev_value:
       count += 1
@@ -2461,19 +2461,19 @@ def parse_select(s):
     m = range_pat.match(s, start)
     if m is not None:
       # Selected from 'left' to (and including) 'right'
-      # Convert into xrange fields, starting from 0
+      # Convert into range fields, starting from 0
       left = int(m.group(1))
       right = m.group(2)
       if not right:
         ranges.append(starting_from(left - 1))
       else:
-        ranges.append(xrange(left - 1, int(right)))
+        ranges.append(range(left - 1, int(right)))
     else:
       # Selected a single value
       m = value_pat.match(s, start)
       if m is not None:
         val = int(m.group(1))
-        ranges.append(xrange(val - 1, val))
+        ranges.append(range(val - 1, val))
       else:
         raise argparse.ArgumentTypeError("Unknown character at position %d of %r" % (start + 1, s))
     start = m.end()

rdkit/Dbase/DbConnection.py

@@ -66,7 +66,7 @@ class DbConnect(object):
     self.user = self.userEntry.GetValue()
     self.password = self.passwdEntry.GetValue()
     self.dbName = self.dbBrowseButton.GetValue()
-    for i in xrange(self.dbTableChoice.Number()):
+    for i in range(self.dbTableChoice.Number()):
       self.dbTableChoice.Delete(0)
 
     names = self.GetTableNames()

rdkit/sping/PDF/pdfmetrics.py

@@ -251,12 +251,12 @@ def parseAFMfile(filename):
   metriclines = []
   between = 0
   for line in alllines:
-    if string.find(string.lower(line), 'endcharmetrics') > -1:
+    if string.find(line.lower(), 'endcharmetrics') > -1:
       between = 0
       break
     if between:
       metriclines.append(line)
-    if string.find(string.lower(line), 'startcharmetrics') > -1:
+    if string.find(line.lower(), 'startcharmetrics') > -1:
       between = 1
 
     # break up - very shaky assumption about array size
@@ -307,7 +307,7 @@ class FontCache:
         return self.getfont('courier')
 
   def stringwidth(self, text, font):
-    widths = self.getfont(string.lower(font))
+    widths = self.getfont(font.lower())
     w = 0
     for char in text:
       w = w + widths[ord(char)]

rdkit/sping/PDF/pidPDF.py

@@ -221,7 +221,7 @@ class PDFCanvas(Canvas):
     if not font.face:
       face = 'serif'
     else:
-      face = string.lower(font.face)
+      face = font.face.lower()
     while face in font_face_map:
       face = font_face_map[face]
     #step 2, - resolve bold/italic to get the right PS font name

rdkit/sping/PS/psmetrics.py

@@ -219,10 +219,10 @@ _Widths = {'StandardEncoding': _stdenc_widths, 'Latin1Encoding': latin1MetricsCa
 
 def stringwidth(text, font, encoding):
   if font in fontinfo.NonRomanFonts:
-    widths = _Widths['StandardEncoding'][string.lower(font)]
+    widths = _Widths['StandardEncoding'][font.lower()]
   else:
     try:
-      widths = _Widths[encoding][string.lower(font)]
+      widths = _Widths[encoding][font.lower()]
     except Exception:
       raise KeyError("Improper encoding {0} or font name {1}".format(encoding, font))
   w = 0

rdkit/sping/Pyart/Fontmapping.py

@@ -74,7 +74,7 @@ def getPyartName(pidfont):
   face = pidfont.face or DefaultFace
   # print "pidfont.face = %s" % pidfont.face
 
-  face = string.lower(face)
+  face = face.lower()
   if face in PidLegalFonts:
     return MapPid2PyartFontName[(PidLegalFonts[face], shape)]
   else:
@@ -93,7 +93,7 @@ getPdfName = getPyartName
 #      face = pidfont.face or DefaultFace
 #      # print "pidfont.face = %s" % pidfont.face
 
-#      face = string.lower(face)
+#      face = face.lower()
 #      if face in PidLegalFonts:
 #          return PDFFontMapping[ ( PidLegalFonts[face], shape) ]
 #      else:

rdkit/sping/Pyart/pidPyart.py

@@ -124,7 +124,7 @@ class PyartCanvas(Canvas):
     }
 
     try:
-      face = piddle_font_map[string.lower(font.face)]
+      face = piddle_font_map[font.face.lower()]
     except Exception:
       return 'Helvetica'
 

rdkit/sping/SVG/pidSVG.py

@@ -74,7 +74,7 @@ def _PointListToSVG(points, dupFirst=0):
 
   """
   outStr = ''
-  for i in xrange(len(points)):
+  for i in range(len(points)):
     outStr = outStr + '%.2f,%.2f ' % (points[i][0], points[i][1])
   # add back on the first point.  This is not required in the spec,
   #  but Adobe's beta-quality viewer seems to not like it being skipped
@@ -130,9 +130,9 @@ class SVGCanvas(Canvas):
     }
 
     try:
-      face = piddle_font_map[string.lower(font.face)]
+      face = piddle_font_map[font.face.lower()]
     except Exception:
-      return piddle_font_map[string.lower('sansserif')]
+      return piddle_font_map['sansserif']
 
     name = face + '-'
     if font.bold and face in ['Courier', 'Helvetica', 'Times']:
@@ -170,7 +170,7 @@ class SVGCanvas(Canvas):
         familyStr = '\'%s\'' % (face)
       else:
         familyStr = face
-      for i in xrange(1, len(font.face)):
+      for i in range(1, len(font.face)):
         face = font.face[i]
         if len(string.split(face)) > 1:
           familyStr = ', \'%s\'' % (face)
@@ -609,7 +609,7 @@ class SVGCanvas(Canvas):
       into SVG is to read it from a file.  So we'll save out to a PNG
       file, then set a link to that in the SVG.
     """
-    imageFileName = '%s-%d.%s' % (self.name, self._nImages, string.lower(self._imageFormat))
+    imageFileName = '%s-%d.%s' % (self.name, self._nImages, self._imageFormat.lower())
     self._nImages = self._nImages + 1
     image.save(imageFileName, format=self._imageFormat)
 

rdkit/sping/TK/pidTK.py

@@ -135,7 +135,7 @@ class FontManager:
     if font.face:
       # check if the user specified a generic face type
       # like serif or monospaced. check is case-insenstive.
-      f = string.lower(font.face)
+      f = font.face.lower()
       if f in self.__alt_faces:
         family = self.__alt_faces[f]
       else: