add index for tags-to-documents

Now we can find writer much faster, because we don't have to execute
a query for documents that match the tags. We can just look up the 
documents in the map.
Speedup: 2-4ms -> 0.002-0.01ms

data-store/src/main/java/org/lucares/pdb/datastore/PdbDB.java

@@ -39,4 +39,8 @@ public class PdbDB {
 		return proposer.propose(query, caretIndex);
 	}
 
+	public List<Doc> getByTags(Tags tags) {
+		
+		return dataStore.getByTags(tags);
+	}
 }

data-store/src/main/java/org/lucares/pdb/datastore/internal/DataStore.java

@@ -48,8 +48,10 @@ public class DataStore {
 
 	// to be guarded by itself
 	private final List<Doc> docIdToDoc = new ArrayList<>();
+	
+	private final ConcurrentHashMap<Tags, List<Doc>> tagsToDocs = new ConcurrentHashMap<>();
 
-	private final Map<String, Map<String, IntList>> keyToValueToDocId = new ConcurrentHashMap<>();
+	private final ConcurrentHashMap<String, Map<String, IntList>> keyToValueToDocId = new ConcurrentHashMap<>();
 
 	private final StringCompressor stringCompressor;
 	private final FolderStorage folderStorage;
@@ -81,11 +83,18 @@ public class DataStore {
 	private void cacheTagToFileMapping(final Tags tags, final Path path) {
 
 		final int docId;
+		final Doc newDoc = new Doc(tags, path);
 		synchronized (docIdToDoc) {			
 			docId = docIdToDoc.size();
-			docIdToDoc.add(new Doc(tags, path));
+			docIdToDoc.add(newDoc);
 		}
-
+		
+		tagsToDocs.compute(tags, (t, listOfDocs) -> {
+			final List<Doc> result = listOfDocs != null ? listOfDocs : new ArrayList<>(2);
+			result.add(newDoc);
+			return result;
+		});
+		
 		for (final String key : tags.getKeys()) {
 			final Map<String, IntList> valueToDocIds = keyToValueToDocId.computeIfAbsent(key, k -> new ConcurrentHashMap<>());
 
@@ -280,4 +289,9 @@ public class DataStore {
 		}
 		return result;
 	}
+
+	public List<Doc> getByTags(Tags tags) {
+		final List<Doc> result = tagsToDocs.getOrDefault(tags, new ArrayList<>(0));
+		return result;
+	}
 }