Algorithm optimizations: fix critical bounds bug, O(1) lookups, faster init

- CRITICAL: Fix EBIMatrix bounds check (i <= rows → i < rows) preventing
  silent out-of-bounds access on matrix edge indices
- EBIMatrix: Use Arrays.fill() for initMatrix (native vs manual loop)
- GraphMatcher: Replace O(n) linear atom-by-ID scan with O(1) HashMap
  lookup — builds ID→Atom maps once, then O(1) per mapping pair
- GraphMatcher: Scale thread pool cap from 4 to 8 for multi-core systems
- Holder: Combine setFingerprint() + setMolMapping() into single loop,
  pre-cache fingerprints to avoid redundant map lookups

All 40 tests pass.

Author: Syed Asad Rahman <asad.rahman@bioinceptionlabs.com>

Author: asad

src/main/java/com/bioinceptionlabs/reactionblast/mapping/algorithm/Holder.java

@@ -99,10 +99,8 @@ public Holder(
         this.structureInformation = reactionContainer;
         this.bestMatchContainer = bestMatchContainer;
         this.hydFPFree = hydFPFree;
-        LOGGER.debug("setFingerprint");
-        setFingerprint();
-        LOGGER.debug("setMolMapping");
-        setMolMapping();
+        LOGGER.debug("setFingerprintAndMolMapping");
+        setFingerprintAndMolMapping();
     }
 
     /**
@@ -171,32 +169,40 @@ private void initialize() {
         carbonOverlapMatrix.initMatrix(0.0);
     }
 
-    private void setFingerprint() {
-        for (int i = 0; i < eductCounter.size(); i++) {
-            for (int j = 0; j < productCounter.size(); j++) {
-                try {
-                    String eductName = eductCounter.get(i).trim();
-                    String productName = productCounter.get(j).trim();
-                    BitSet hydrogenEductFP = hydFPFree.getFingerPrint(eductName);
-                    BitSet hydrogenProductFP = hydFPFree.getFingerPrint(productName);
-                    float hydrogenSimVal = getTanimotoSimilarity(hydrogenEductFP, hydrogenProductFP);
-                    LOGGER.debug("FP " + hydrogenSimVal);
-                    fpSimMatrixWithoutHydrogen.setValue(i, j, hydrogenSimVal);
-                } catch (Exception ex) {
-                    LOGGER.error(SEVERE, null, ex);
-                }
+    private void setFingerprintAndMolMapping() {
+        // Pre-cache trimmed names and fingerprints to avoid redundant lookups
+        int eSize = eductCounter.size();
+        int pSize = productCounter.size();
+        String[] eNames = new String[eSize];
+        BitSet[] eFPs = new BitSet[eSize];
+        for (int i = 0; i < eSize; i++) {
+            eNames[i] = eductCounter.get(i).trim();
+            try {
+                eFPs[i] = hydFPFree.getFingerPrint(eNames[i]);
+            } catch (Exception ex) {
+                LOGGER.error(SEVERE, null, ex);
             }
         }
-    }
-
-    private void setMolMapping() {
-        for (int i = 0; i < eductCounter.size(); i++) {
-            for (int j = 0; j < productCounter.size(); j++) {
+        String[] pNames = new String[pSize];
+        BitSet[] pFPs = new BitSet[pSize];
+        for (int j = 0; j < pSize; j++) {
+            pNames[j] = productCounter.get(j).trim();
+            try {
+                pFPs[j] = hydFPFree.getFingerPrint(pNames[j]);
+            } catch (Exception ex) {
+                LOGGER.error(SEVERE, null, ex);
+            }
+        }
+        // Single combined loop for fingerprint similarity + mol mapping
+        for (int i = 0; i < eSize; i++) {
+            for (int j = 0; j < pSize; j++) {
                 try {
-                    String eductName = eductCounter.get(i).trim();
-                    String productName = productCounter.get(j).trim();
-                    MolMapping m = new MolMapping(eductName, productName, i, j);
-                    getMappingMolPair().add(m);
+                    if (eFPs[i] != null && pFPs[j] != null) {
+                        float hydrogenSimVal = getTanimotoSimilarity(eFPs[i], pFPs[j]);
+                        LOGGER.debug("FP " + hydrogenSimVal);
+                        fpSimMatrixWithoutHydrogen.setValue(i, j, hydrogenSimVal);
+                    }
+                    getMappingMolPair().add(new MolMapping(eNames[i], pNames[j], i, j));
                 } catch (Exception ex) {
                     LOGGER.error(SEVERE, null, ex);
                 }

src/main/java/com/bioinceptionlabs/reactionblast/mapping/graph/GraphMatcher.java

@@ -154,8 +154,8 @@ public synchronized static Collection<MCSSolution> matcher(Holder mh) throws Exc
                 threadsAvailable = jobMap.size();
             }
 
-            if (threadsAvailable > 4) {
-                threadsAvailable = 4;
+            if (threadsAvailable > 8) {
+                threadsAvailable = 8;
             }
 
             LOGGER.debug(threadsAvailable + " threads requested for MCS in " + mh.getTheory());
@@ -385,13 +385,24 @@ static MCSSolution replicateMappingOnContainers(Holder mh, Combination solution,
 
             AtomAtomMapping atomAtomMapping = mcs.getAtomAtomMapping();
             AtomAtomMapping atomAtomMappingNew = new AtomAtomMapping(q, t);
-            atomAtomMapping.getMappingsByAtoms().keySet().stream().forEach((a) -> {
-                IAtom atomByID1 = getAtomByID(q, a);
-                IAtom b = atomAtomMapping.getMappingsByAtoms().get(a);
-                IAtom atomByID2 = getAtomByID(t, b);
-//                if (DEBUG) {
-//                    out.println("atomByID1 " + atomByID1.getID() + " atomByID2 " + atomByID2.getID());
-//                }
+
+            // Build ID→Atom maps for O(1) lookup instead of O(n) linear scan
+            Map<String, IAtom> qIdMap = new java.util.HashMap<>();
+            for (IAtom a : q.atoms()) {
+                if (a.getID() != null) {
+                    qIdMap.put(a.getID(), a);
+                }
+            }
+            Map<String, IAtom> tIdMap = new java.util.HashMap<>();
+            for (IAtom a : t.atoms()) {
+                if (a.getID() != null) {
+                    tIdMap.put(a.getID(), a);
+                }
+            }
+
+            atomAtomMapping.getMappingsByAtoms().forEach((a, b) -> {
+                IAtom atomByID1 = a.getID() != null ? qIdMap.get(a.getID()) : null;
+                IAtom atomByID2 = b.getID() != null ? tIdMap.get(b.getID()) : null;
                 if (atomByID1 != null && atomByID2 != null) {
                     atomAtomMappingNew.put(atomByID1, atomByID2);
                 } else {
@@ -410,15 +421,4 @@ static MCSSolution replicateMappingOnContainers(Holder mh, Combination solution,
         return null;
     }
 
-    private static IAtom getAtomByID(IAtomContainer ac, IAtom atom) {
-        if (atom.getID() == null) {
-            return null;
-        }
-        for (IAtom a : ac.atoms()) {
-            if (atom.getID().equals(a.getID())) {
-                return a;
-            }
-        }
-        return null;
-    }
 }

src/main/java/com/bioinceptionlabs/reactionblast/tools/EBIMatrix.java

@@ -410,9 +410,7 @@ public EBIMatrix(double vals[], int m) {
      */
     public synchronized void initMatrix(double v) {
         for (int i = 0; i < rows; i++) {
-            for (int j = 0; j < columns; j++) {
-                matrix[i][j] = v;
-            }
+            java.util.Arrays.fill(matrix[i], v);
         }
     }
 
@@ -427,11 +425,10 @@ public synchronized void initMatrix(double v) {
     public synchronized double getValue(int i, int j) {
 
         double val = -1.0d;
-        if (i <= rows && j <= columns) {
+        if (i >= 0 && i < rows && j >= 0 && j < columns) {
             val = matrix[i][j];
         } else {
-
-            LOGGER.debug("Error: Array of out bound");
+            LOGGER.debug("Error: Array out of bounds [" + i + "," + j + "] for [" + rows + "," + columns + "]");
         }
         return val;
     }
@@ -564,7 +561,7 @@ public synchronized boolean setValue(int row, int col, double value) {
         double val = value;
         boolean flag = false;
 
-        if (row <= rows && col <= columns) {
+        if (row >= 0 && row < rows && col >= 0 && col < columns) {
             flag = true;
             matrix[row][col] = val;
         } else {