From 72768357cb5d47a331c1917f47d16d3a4f0eee09 Mon Sep 17 00:00:00 2001
From: lifulong <lifulong@kkworld.com>
Date: Tue, 31 Dec 2019 16:16:55 +0800
Subject: [PATCH 1/2] real time update words dict support

---
 .../huaban/analysis/jieba/JiebaSegmenter.java |  2 +-
 .../huaban/analysis/jieba/WordDictionary.java | 32 ++++++++++++++-----
 .../analysis/jieba/viterbi/FinalSeg.java      |  2 +-
 .../analysis/jieba/JiebaSegmenterTest.java    |  2 +-
 4 files changed, 27 insertions(+), 11 deletions(-)

diff --git a/src/main/java/com/huaban/analysis/jieba/JiebaSegmenter.java b/src/main/java/com/huaban/analysis/jieba/JiebaSegmenter.java
index 9010f1f..915c031 100644
--- a/src/main/java/com/huaban/analysis/jieba/JiebaSegmenter.java
+++ b/src/main/java/com/huaban/analysis/jieba/JiebaSegmenter.java
@@ -1,5 +1,6 @@
 package com.huaban.analysis.jieba;
 
+import java.nio.file.Path;
 import java.util.ArrayList;
 import java.util.HashMap;
 import java.util.List;
@@ -28,7 +29,6 @@ public void initUserDict(Path path){
 
     public void initUserDict(String[] paths){
         wordDict.init(paths);
-
     }
     
     private Map<Integer, List<Integer>> createDAG(String sentence) {
diff --git a/src/main/java/com/huaban/analysis/jieba/WordDictionary.java b/src/main/java/com/huaban/analysis/jieba/WordDictionary.java
index da40753..8823584 100644
--- a/src/main/java/com/huaban/analysis/jieba/WordDictionary.java
+++ b/src/main/java/com/huaban/analysis/jieba/WordDictionary.java
@@ -9,12 +9,8 @@
 import java.io.InputStreamReader;
 import java.nio.charset.Charset;
 import java.nio.charset.StandardCharsets;
-import java.util.HashMap;
-import java.util.HashSet;
-import java.util.Locale;
-import java.util.Map;
+import java.util.*;
 import java.util.Map.Entry;
-import java.util.Set;
 
 
 public class WordDictionary {
@@ -145,7 +141,7 @@ public void loadDict() {
     }
 
 
-    private String addWord(String word) {
+    public String addWord(String word) {
         if (null != word && !"".equals(word.trim())) {
             String key = word.trim().toLowerCase(Locale.getDefault());
             _dict.fillSegment(key.toCharArray());
@@ -154,7 +150,27 @@ private String addWord(String word) {
         else
             return null;
     }
-
+    
+    
+    public String addWord(String word, double freq) {
+        word = addWord(word);
+        freqs.put(word, freq);
+        return word;
+    }
+    
+    
+    public void addWords(List<String> words) {
+        for(String word : words)
+            addWord(word);
+    }
+    
+    
+    public void addWords(Map<String, Double> words) {
+        for(Map.Entry<String, Double> entry : words.entrySet()) {
+            addWord(entry.getKey(), entry.getValue());
+        }
+    }
+    
 
     public void loadUserDict(Path userDict) {
         loadUserDict(userDict, StandardCharsets.UTF_8);
@@ -183,7 +199,7 @@ public void loadUserDict(Path userDict, Charset charset) {
                 double freq = 3.0d;
                 if (tokens.length == 2)
                     freq = Double.valueOf(tokens[1]);
-                word = addWord(word); 
+                word = addWord(word);
                 freqs.put(word, Math.log(freq / total));
                 count++;
             }
diff --git a/src/main/java/com/huaban/analysis/jieba/viterbi/FinalSeg.java b/src/main/java/com/huaban/analysis/jieba/viterbi/FinalSeg.java
index 8a79eb5..5aa8b83 100644
--- a/src/main/java/com/huaban/analysis/jieba/viterbi/FinalSeg.java
+++ b/src/main/java/com/huaban/analysis/jieba/viterbi/FinalSeg.java
@@ -26,7 +26,7 @@ public class FinalSeg {
     private static Map<Character, Double> start;
     private static Map<Character, Map<Character, Double>> trans;
     private static Map<Character, char[]> prevStatus;
-    private static Double MIN_FLOAT = -3.14e100;;
+    private static Double MIN_FLOAT = -3.14e100;
 
 
     private FinalSeg() {
diff --git a/src/test/java/com/huaban/analysis/jieba/JiebaSegmenterTest.java b/src/test/java/com/huaban/analysis/jieba/JiebaSegmenterTest.java
index b31f98b..f7c9129 100644
--- a/src/test/java/com/huaban/analysis/jieba/JiebaSegmenterTest.java
+++ b/src/test/java/com/huaban/analysis/jieba/JiebaSegmenterTest.java
@@ -11,6 +11,7 @@
 
 import junit.framework.TestCase;
 
+import org.junit.Ignore;
 import org.junit.Test;
 
 import com.huaban.analysis.jieba.JiebaSegmenter.SegMode;
@@ -163,7 +164,6 @@ public void testBugSentence() {
         }
     }
 
-
     @Test
     public void testSegmentSpeed() {
         long length = 0L;

From 1d55902e2eb8f5b7831f1b51c19e95c0bb3a1f34 Mon Sep 17 00:00:00 2001
From: lifulong <lifulong@kkworld.com>
Date: Tue, 31 Dec 2019 16:19:29 +0800
Subject: [PATCH 2/2] unset unuseful update

---
 .../java/com/huaban/analysis/jieba/JiebaSegmenterTest.java    | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/src/test/java/com/huaban/analysis/jieba/JiebaSegmenterTest.java b/src/test/java/com/huaban/analysis/jieba/JiebaSegmenterTest.java
index f7c9129..4de6692 100644
--- a/src/test/java/com/huaban/analysis/jieba/JiebaSegmenterTest.java
+++ b/src/test/java/com/huaban/analysis/jieba/JiebaSegmenterTest.java
@@ -11,7 +11,6 @@
 
 import junit.framework.TestCase;
 
-import org.junit.Ignore;
 import org.junit.Test;
 
 import com.huaban.analysis.jieba.JiebaSegmenter.SegMode;
@@ -179,8 +178,7 @@ public void testSegmentSpeed() {
         System.out.println(String.format(Locale.getDefault(), "time elapsed:%d, rate:%fkb/s, sentences:%.2f/s", elapsed,
             (length * 1.0) / 1024.0f / (elapsed * 1.0 / 1000.0f), wordCount * 1000.0f / (elapsed * 1.0)));
     }
-
-
+    
     @Test
     public void testLongTextSegmentSpeed() {
         long length = 0L;