Work around limitation of absolute SMILES with unknown elements. We e…

…ncode the unknown element as Rf (Rutherfordium - the highest supported by InChI v1.3 - JNI InChI current). We can still encode structures with Rf providing there are no unknown atoms.
cdk · Mar 6, 2017 · cc4da2b · cc4da2b
1 parent 31ae0dc
commit cc4da2b
Show file tree

Hide file tree

Showing 2 changed files with 42 additions and 0 deletions.
diff --git a/storage/inchi/src/test/java/org/openscience/cdk/smiles/AbsoluteSMILESTest.java b/storage/inchi/src/test/java/org/openscience/cdk/smiles/AbsoluteSMILESTest.java
@@ -27,6 +27,7 @@
 import com.google.common.base.Joiner;
 import org.junit.Assert;
 import org.junit.Test;
+import org.openscience.cdk.exception.CDKException;
 import org.openscience.cdk.interfaces.IAtomContainer;
 import org.openscience.cdk.silent.SilentChemObjectBuilder;
 
@@ -112,6 +113,21 @@ public void dbStereoCanonGeneration() throws Exception {
                             cansmi.create(smipar.parseSmiles(cansmi.create(mol))));
     }
 
+    @Test
+    public void smilesWithUnknownElem() throws Exception {
+        test("*CC", "CC*");
+    }
+
+    @Test
+    public void rfElement() throws Exception {
+        test("[Rf]");
+    }
+
+    @Test(expected = CDKException.class)
+    public void problematic() throws Exception {
+        test("*[Rf]");
+    }
+
     static void test(String... inputs) throws Exception {
 
         SmilesParser sp = new SmilesParser(SilentChemObjectBuilder.getInstance());

diff --git a/storage/smiles/src/main/java/org/openscience/cdk/smiles/SmilesGenerator.java b/storage/smiles/src/main/java/org/openscience/cdk/smiles/SmilesGenerator.java
@@ -23,6 +23,7 @@
 package org.openscience.cdk.smiles;
 
 import org.openscience.cdk.CDKConstants;
+import org.openscience.cdk.config.Elements;
 import org.openscience.cdk.exception.CDKException;
 import org.openscience.cdk.graph.ConnectedComponents;
 import org.openscience.cdk.graph.GraphUtil;
@@ -47,6 +48,7 @@
 import java.lang.reflect.Method;
 import java.util.ArrayList;
 import java.util.Collection;
+import java.util.Collections;
 import java.util.HashMap;
 import java.util.HashSet;
 import java.util.IdentityHashMap;
@@ -690,6 +692,13 @@ private static long[] inchiNumbers(IAtomContainer container) throws CDKException
         // class each time
         String cname = "org.openscience.cdk.graph.invariant.InChINumbersTools";
         String mname = "getUSmilesNumbers";
+
+        List<IAtom> rgrps = getRgrps(container, Elements.Rutherfordium);
+        for (IAtom rgrp : rgrps) {
+            rgrp.setAtomicNumber(Elements.Rutherfordium.number());
+            rgrp.setSymbol(Elements.Rutherfordium.symbol());
+        }
+
         try {
             Class<?> c = Class.forName(cname);
             Method method = c.getDeclaredMethod("getUSmilesNumbers", IAtomContainer.class);
@@ -703,7 +712,24 @@ private static long[] inchiNumbers(IAtomContainer container) throws CDKException
             throw new CDKException("An InChI could not be generated and used to canonise SMILES: " + e.getMessage(), e);
         } catch (IllegalAccessException e) {
             throw new CDKException("Could not access method to obtain InChI numbers.");
+        } finally {
+            for (IAtom rgrp : rgrps) {
+                rgrp.setAtomicNumber(Elements.Unknown.number());
+                rgrp.setSymbol("*");
+            }
+        }
+    }
+
+    private static List<IAtom> getRgrps(IAtomContainer container, Elements reversed) {
+        List<IAtom> res = new ArrayList<>();
+        for (IAtom atom : container.atoms()) {
+            if (atom.getAtomicNumber() == 0) {
+                res.add(atom);
+            } else if (atom.getAtomicNumber() == reversed.number()) {
+                return Collections.emptyList();
+            }
         }
+        return res;
     }
 
     // utility safety check to guard against invalid state