集群环境中CoreStopWordDictionary适配IOAdapter： #530

hankcs · May 16, 2017 · 06686dd · 06686dd
1 parent ab1c2c3
commit 06686dd
Show file tree

Hide file tree

Showing 4 changed files with 36 additions and 7 deletions.
diff --git a/src/main/java/com/hankcs/hanlp/collection/MDAG/MDAG.java b/src/main/java/com/hankcs/hanlp/collection/MDAG/MDAG.java
@@ -25,8 +25,11 @@
 
 import com.hankcs.hanlp.corpus.io.ByteArray;
 import com.hankcs.hanlp.corpus.io.ICacheAble;
-import static com.hankcs.hanlp.HanLP.Config.IOAdapter;
-import java.io.*;
+import com.hankcs.hanlp.corpus.io.IOUtil;
+import java.io.BufferedReader;
+import java.io.DataOutputStream;
+import java.io.File;
+import java.io.IOException;
 import java.util.*;
 import java.util.Map.Entry;
 
@@ -157,6 +160,16 @@ public boolean satisfiesCondition(String str1, String str2)
     }
     /////
 
+    /**
+     * 从文件路径构造
+     * @param path
+     * @throws IOException
+     */
+    public MDAG(String path) throws IOException
+    {
+        this(IOUtil.newBufferedReader(path));
+    }
+
 
     /**
      * 从一个文件建立MDAG<br>
@@ -168,10 +181,16 @@ public boolean satisfiesCondition(String str1, String str2)
      */
     public MDAG(File dataFile) throws IOException
     {
-        BufferedReader dataFileBufferedReader = new BufferedReader(new InputStreamReader(IOAdapter == null ?
-                                                                                                 new FileInputStream(dataFile) :
-                IOAdapter.open(dataFile.getAbsolutePath())
-                , "UTF-8"));
+        this(IOUtil.newBufferedReader(dataFile.getPath()));
+    }
+
+    /**
+     * 从一个打开的BufferedReader构造
+     * @param dataFileBufferedReader
+     * @throws IOException
+     */
+    public MDAG(BufferedReader dataFileBufferedReader) throws IOException
+    {
         String currentString = "";
         String previousString = "";
 

diff --git a/src/main/java/com/hankcs/hanlp/collection/MDAG/MDAGSet.java b/src/main/java/com/hankcs/hanlp/collection/MDAG/MDAGSet.java
@@ -37,6 +37,11 @@ public MDAGSet()
     {
     }
 
+    public MDAGSet(String dictionaryPath) throws IOException
+    {
+        super(dictionaryPath);
+    }
+
     @Override
     public int size()
     {

diff --git a/src/main/java/com/hankcs/hanlp/dictionary/stopword/CoreStopWordDictionary.java b/src/main/java/com/hankcs/hanlp/dictionary/stopword/CoreStopWordDictionary.java
@@ -39,7 +39,7 @@ public class CoreStopWordDictionary
         {
             try
             {
-                dictionary = new StopWordDictionary(new File(HanLP.Config.CoreStopWordDictionaryPath));
+                dictionary = new StopWordDictionary(HanLP.Config.CoreStopWordDictionaryPath);
                 DataOutputStream out = new DataOutputStream(IOUtil.newOutputStream(HanLP.Config.CoreStopWordDictionaryPath + Predefine.BIN_EXT));
                 dictionary.save(out);
                 out.close();

diff --git a/src/main/java/com/hankcs/hanlp/dictionary/stopword/StopWordDictionary.java b/src/main/java/com/hankcs/hanlp/dictionary/stopword/StopWordDictionary.java
@@ -39,6 +39,11 @@ public StopWordDictionary()
     {
     }
 
+    public StopWordDictionary(String stopWordDictionaryPath) throws IOException
+    {
+        super(stopWordDictionaryPath);
+    }
+
     @Override
     public boolean shouldInclude(Term term)
     {