Merge pull request #2471 from esaunders/snippet_fix

Snippet fix

KeywordSearch/src/org/sleuthkit/autopsy/keywordsearch/KeywordHit.java

@@ -30,7 +30,7 @@ import org.sleuthkit.datamodel.TskCoreException;
  * keyword was found and the file available to clients. Artifact keyword hits
  * also make the artifact available to clients.
  */
-class KeywordHit {
+class KeywordHit implements Comparable<KeywordHit> {
 
     private final String solrDocumentId;
     private final long solrObjectId;
@@ -140,4 +140,23 @@ class KeywordHit {
         return hash;
     }
 
+    @Override
+    public int compareTo(KeywordHit o) {
+        if (this.solrObjectId < o.solrObjectId) {
+            // Out object id is less than the other object id
+            return -1;
+        } else if (this.solrObjectId == o.solrObjectId) {
+            // Hits have same object id
+            if (this.chunkId < o.chunkId) {
+                // Our chunk id is lower than the other chunk id
+                return -1;
+            } else {
+                // Our chunk id is either greater than or equal to the other chunk id
+                return this.chunkId == o.chunkId ? 0 : 1;
+            }
+        } else {
+            // Our object id is greater than the other object id
+            return 1;
+        }
+    }
 }

KeywordSearch/src/org/sleuthkit/autopsy/keywordsearch/QueryResults.java

@@ -131,7 +131,7 @@ class QueryResults {
                 subProgress.progress(keywordList.getName() + ": " + hitDisplayStr, unitProgress);
             }
 
-            for (KeywordHit hit : getOneHitPerObject(keyword)) {
+            for (KeywordHit hit : getResults(keyword)) {
                 String termString = keyword.getSearchTerm();
                 final String snippetQuery = KeywordSearchUtil.escapeLuceneQuery(termString);
                 String snippet;
@@ -174,29 +174,6 @@ class QueryResults {
         return newArtifacts;
     }
 
-    /**
-     * Gets the first hit of the keyword.
-     *
-     * @param keyword
-     *
-     * @return Collection<KeywordHit> containing KeywordHits with lowest
-     *         SolrObjectID-ChunkID pairs.
-     */
-    private Collection<KeywordHit> getOneHitPerObject(Keyword keyword) {
-
-        HashMap<Long, KeywordHit> hits = new HashMap<>();
-
-        // create a list of KeywordHits. KeywordHits with lowest chunkID is added the the list.
-        for (KeywordHit hit : getResults(keyword)) {
-            if (!hits.containsKey(hit.getSolrObjectId())) {
-                hits.put(hit.getSolrObjectId(), hit);
-            } else if (hit.getChunkId() < hits.get(hit.getSolrObjectId()).getChunkId()) {
-                hits.put(hit.getSolrObjectId(), hit);
-            }
-        }
-        return hits.values();
-    }
-
     /**
      * Generate an ingest inbox message for given keyword in given file
      *

KeywordSearch/src/org/sleuthkit/autopsy/keywordsearch/SearchRunner.java

@@ -20,6 +20,7 @@ package org.sleuthkit.autopsy.keywordsearch;
 
 import java.util.ArrayList;
 import java.util.Collection;
+import java.util.Collections;
 import java.util.HashMap;
 import java.util.HashSet;
 import java.util.List;
@@ -580,8 +581,9 @@ public final class SearchRunner {
          * previously seen a hit for the keyword.
          * 
          * @param queryResult The results returned by a keyword search.
-         * @return The set of hits found by the most recent search for objects
+         * @return A unique set of hits found by the most recent search for objects
-         * that have not previously had a hit.
+         * that have not previously had a hit. The hits will be for the lowest
+         * numbered chunk associated with the object.
          *
          */
         private QueryResults filterResults(QueryResults queryResult) {
@@ -596,6 +598,10 @@ public final class SearchRunner {
                 // This may well include duplicates of hits we've seen in earlier periodic searches.
                 List<KeywordHit> queryTermResults = queryResult.getResults(keyword);
 
+                // Sort the hits for this keyword so that we are always 
+                // guaranteed to return the hit for the lowest chunk.
+                Collections.sort(queryTermResults);
+
                 // This will be used to build up the hits we haven't seen before
                 // for this keyword.
                 List<KeywordHit> newUniqueHits = new ArrayList<>();