Issue #2 implemented whole word matching

README.md

@@ -79,6 +79,19 @@ conflict resolution rules: 1) longer matches prevail over shorter matches, 2) le
 There is only one result now:
 * "hot chocolate" starting at position 0, ending at position 12
 
+If you want the algorithm to only check for whole words, you can tell the Trie to do so:
+
+```java
+    Trie trie = new Trie().onlyWholeWords();
+    trie.addKeyword("sugar");
+    Collection<Emit> emits = trie.parseText("sugarcane sugarcane sugar canesugar");
+```
+
+In this case, it will only find one match, whereas it would normally find four. The sugarcane/canesugar words
+are discarded because they are partial matches.
+
+
+
 License
 -------
    Licensed under the Apache License, Version 2.0 (the "License");

src/main/java/org/ahocorasick/trie/Trie.java

@@ -36,6 +36,11 @@ public class Trie {
         return this;
     }
 
+    public Trie onlyWholeWords() {
+        this.trieConfig.setOnlyWholeWords(true);
+        return this;
+    }
+
     public void addKeyword(String keyword) {
 
         State currentState = this.rootState;
@@ -58,6 +63,10 @@ public class Trie {
             position++;
         }
 
+        if (trieConfig.isOnlyWholeWords()) {
+            removePartialMatches(text, collectedEmits);
+        }
+
         if (!trieConfig.isAllowOverlaps()) {
             IntervalTree intervalTree = new IntervalTree((List<Intervalable>)(List<?>)collectedEmits);
             intervalTree.removeOverlaps((List<Intervalable>) (List<?>) collectedEmits);
@@ -66,6 +75,24 @@ public class Trie {
         return collectedEmits;
     }
 
+    private void removePartialMatches(String searchText, List<Emit> collectedEmits) {
+        long size = searchText.length();
+        List<Emit> removeEmits = new ArrayList<Emit>();
+        for (Emit emit : collectedEmits) {
+            if ((emit.getStart() == 0 ||
+                 searchText.charAt(emit.getStart() - 1) == ' ') &&
+                (emit.getEnd() == size ||
+                 searchText.charAt(emit.getEnd() + 1) == ' ')) {
+                continue;
+            }
+            removeEmits.add(emit);
+        }
+
+        for (Emit removeEmit : removeEmits) {
+            collectedEmits.remove(removeEmit);
+        }
+    }
+
     private State getState(State currentState, Character character) {
         State newCurrentState = currentState.nextState(character);
         while (newCurrentState == null) {

src/main/java/org/ahocorasick/trie/TrieConfig.java

@@ -4,6 +4,8 @@ public class TrieConfig {
 
     private boolean allowOverlaps = true;
 
+    private boolean onlyWholeWords = false;
+
     public boolean isAllowOverlaps() {
         return allowOverlaps;
     }
@@ -12,4 +14,12 @@ public class TrieConfig {
         this.allowOverlaps = allowOverlaps;
     }
 
+    public boolean isOnlyWholeWords() {
+        return onlyWholeWords;
+    }
+
+    public void setOnlyWholeWords(boolean onlyWholeWords) {
+        this.onlyWholeWords = onlyWholeWords;
+    }
+
 }

src/test/java/org/ahocorasick/trie/TrieTest.java

@@ -123,12 +123,18 @@ public class TrieTest {
         trie.addKeyword("n");
         trie.addKeyword("urning");
         Collection<Emit> emits = trie.parseText("Turning");
-        for (Emit emit : emits) {
-            System.out.println(emit.getStart()+":"+emit.getEnd()+"="+emit.getKeyword());
-        }
         assertEquals(2, emits.size());
     }
 
+    @Test
+    public void partialMatch() {
+        Trie trie = new Trie().onlyWholeWords();
+        trie.addKeyword("sugar");
+        Collection<Emit> emits = trie.parseText("sugarcane sugarcane sugar canesugar"); // left, middle, right test
+        assertEquals(1, emits.size()); // Match must not be made
+        checkEmit(emits.iterator().next(), 20, 24, "sugar");
+    }
+
     private void checkEmit(Emit next, int expectedStart, int expectedEnd, String expectedKeyword) {
         assertEquals(expectedStart, next.getStart());
         assertEquals(expectedEnd, next.getEnd());