支持jdk1.6

提高代码可重用性
hankcs · Feb 12, 2015 · bda5129 · bda5129
1 parent 0570952
commit bda5129
Show file tree

Hide file tree

Showing 106 changed files with 579 additions and 649 deletions.
diff --git a/.gitignore b/.gitignore
@@ -14,8 +14,11 @@ hs_err_pid*
 # IDE相关，不再同步
 /.idea
 /*.iml
+/.classpath
+/*.eml
+/.project
 # 以下两项是词典和模型的二进制文件，请不要提交到版本库
 *.dat
 *.bin
 # 我做实验用的目录
-/data/test
+/data/test
diff --git a/data/dictionary/person/nr.tr.txt b/data/dictionary/person/nr.tr.txt
@@ -5,7 +5,7 @@ C,0,0,0,155393,0,0,0,0,0,0,0,47,461,0,0
 D,2409,3427,0,0,0,0,0,0,141568,16990,0,0,0,63,0
 E,1491,2198,0,0,1000,0,0,0,81654,9104,0,4,0,45,0
 F,0,20,0,0,0,0,0,0,0,0,0,0,0,0,0
-G,6,0,0,0,0,0,0,0,2907,1000,0,0,0,0,0
+G,6,0,0,1000,0,0,0,0,2907,1000,0,0,0,0,0
 K,0,217510,0,0,0,18,0,0,0,0,0,0,0,7506,11
 L,256487,0,0,0,0,0,0,8039,0,0,0,7,0,0,0
 M,0,29629,0,0,0,0,0,0,0,0,0,0,0,526,0

diff --git a/data/dictionary/person/nr.txt b/data/dictionary/person/nr.txt
@@ -5821,7 +5821,7 @@ ypy L 1
 囊中 L 1
 囊括 K 1
 囚 L 1
-四 C 136 L 131 D 11 E 7 K 2
+四 C 136 L 131 G 100 D 11 E 7 K 2
 四人 V 2
 四代 Z 1
 四倍 Z 1
@@ -22159,3 +22159,4 @@ ypy L 1
 ＣＥＯ K 4
 Ｒ４４ L 1
 ～ L 1
+汪洋张 A 1
diff --git a/pom.xml b/pom.xml
@@ -4,7 +4,7 @@
 
     <groupId>com.hankcs</groupId>
     <artifactId>HanLP</artifactId>
-    <version>1.0</version>
+    <version>1.0.1</version>
 
     <name>HanLP</name>
     <url>http://www.hankcs.com/</url>
@@ -103,7 +103,7 @@
         </plugins>
     </build>
     <properties>
-        <java.version>1.7</java.version>
+        <java.version>1.6</java.version>
         <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
 
         <maven-compiler-plugin.version>3.0</maven-compiler-plugin.version>

diff --git a/src/main/java/com/hankcs/hanlp/algoritm/BinarySearch.java b/src/main/java/com/hankcs/hanlp/algoritm/BinarySearch.java
@@ -23,11 +23,11 @@ public class BinarySearch
 {
     public static void main(String[] args)
     {
-        TreeSet<Long> setA = new TreeSet<>();
+        TreeSet<Long> setA = new TreeSet<Long>();
         setA.add(5L);
         setA.add(4L);
 
-        TreeSet<Long> setB = new TreeSet<>();
+        TreeSet<Long> setB = new TreeSet<Long>();
         setB.add(1L);
         setB.add(2L);
         setB.add(3L);

diff --git a/src/main/java/com/hankcs/hanlp/algoritm/Dijkstra.java b/src/main/java/com/hankcs/hanlp/algoritm/Dijkstra.java
@@ -29,7 +29,7 @@ public class Dijkstra
 {
     public static List<Vertex> compute(Graph graph)
     {
-        List<Vertex> resultList = new LinkedList<>();
+        List<Vertex> resultList = new LinkedList<Vertex>();
         Vertex[] vertexes = graph.getVertexes();
         List<EdgeFrom>[] edgesTo = graph.getEdgesTo();
         double[] d = new double[vertexes.length];

diff --git a/src/main/java/com/hankcs/hanlp/algoritm/ViterbiEx.java b/src/main/java/com/hankcs/hanlp/algoritm/ViterbiEx.java
@@ -35,7 +35,7 @@ public ViterbiEx(List<EnumItem<E>> roleTagList, TransformMatrixDictionary<E> tra
     public List<E> computeTagList()
     {
         int length = roleTagList.size() - 1;
-        List<E> tagList = new LinkedList<>();
+        List<E> tagList = new LinkedList<E>();
         double[][] cost = new double[length][];
         Iterator<EnumItem<E>> iterator = roleTagList.iterator();
         EnumItem<E> start = iterator.next();

diff --git a/src/main/java/com/hankcs/hanlp/algoritm/ahocorasick/trie/Trie.java b/src/main/java/com/hankcs/hanlp/algoritm/ahocorasick/trie/Trie.java
@@ -163,7 +163,7 @@ public Collection<Emit> parseText(String text)
     private static void remainLongest(Collection<Emit> collectedEmits)
     {
         if (collectedEmits.size() < 2) return;
-        Map<Integer, Emit> emitMapStart = new TreeMap<>();
+        Map<Integer, Emit> emitMapStart = new TreeMap<Integer, Emit>();
         for (Emit emit : collectedEmits)
         {
             Emit pre = emitMapStart.get(emit.getStart());
@@ -178,7 +178,7 @@ private static void remainLongest(Collection<Emit> collectedEmits)
             collectedEmits.addAll(emitMapStart.values());
             return;
         }
-        Map<Integer, Emit> emitMapEnd = new TreeMap<>();
+        Map<Integer, Emit> emitMapEnd = new TreeMap<Integer, Emit>();
         for (Emit emit : emitMapStart.values())
         {
             Emit pre = emitMapEnd.get(emit.getEnd());

diff --git a/src/main/java/com/hankcs/hanlp/collection/AhoCorasick/AhoCorasickDoubleArrayTrie.java b/src/main/java/com/hankcs/hanlp/collection/AhoCorasick/AhoCorasickDoubleArrayTrie.java
@@ -69,7 +69,7 @@ public List<Hit<V>> parseText(String text)
     {
         int position = 1;
         int currentState = 0;
-        List<Hit<V>> collectedEmits = new LinkedList<>();
+        List<Hit<V>> collectedEmits = new LinkedList<Hit<V>>();
         for (int i = 0; i < text.length(); ++i)
         {
             currentState = getState(currentState, text.charAt(i));
@@ -425,7 +425,7 @@ private int fetch(State parent, List<Map.Entry<Integer, State>> siblings)
         {
             State fakeNode = new State(-(parent.getDepth() + 1));  // 此节点是parent的子节点，同时具备parent的输出
             fakeNode.addEmit(parent.getLargestValueId());
-            siblings.add(new AbstractMap.SimpleEntry<>(0, fakeNode));
+            siblings.add(new AbstractMap.SimpleEntry<Integer, State>(0, fakeNode));
         }
         for (Map.Entry<Character, State> entry : parent.getSuccess().entrySet())
         {
@@ -814,7 +814,7 @@ private void buildDoubleArrayTrie(Set<String> keySet)
 
             State root_node = this.rootState;
 
-            List<Map.Entry<Integer, State>> siblings = new ArrayList<>(root_node.getSuccess().entrySet().size());
+            List<Map.Entry<Integer, State>> siblings = new ArrayList<Map.Entry<Integer, State>>(root_node.getSuccess().entrySet().size());
             fetch(root_node, siblings);
             insert(siblings);
         }
@@ -917,7 +917,7 @@ else if (first == 0)
 
             for (Map.Entry<Integer, State> sibling : siblings)
             {
-                List<Map.Entry<Integer, State>> new_siblings = new ArrayList<>(sibling.getValue().getSuccess().entrySet().size() + 1);
+                List<Map.Entry<Integer, State>> new_siblings = new ArrayList<Map.Entry<Integer, State>>(sibling.getValue().getSuccess().entrySet().size() + 1);
 
                 if (fetch(sibling.getValue(), new_siblings) == 0)  // 一个词的终止且不为其他词的前缀，其实就是叶子节点
                 {

diff --git a/src/main/java/com/hankcs/hanlp/collection/MDAG/MDAG.java b/src/main/java/com/hankcs/hanlp/collection/MDAG/MDAG.java
@@ -760,7 +760,7 @@ public void unSimplify()
         if (sourceNode == null)
         {
             sourceNode = new MDAGNode(false);
-            equivalenceClassMDAGNodeHashMap = new HashMap<>();
+            equivalenceClassMDAGNodeHashMap = new HashMap<MDAGNode, MDAGNode>();
             MDAGNode[] toNodeArray = new MDAGNode[mdagDataArray.length];
             createMDAGNode(simplifiedSourceNode, -1, toNodeArray, new MDAGNode[mdagDataArray.length]);
             // 构建注册表
@@ -898,7 +898,7 @@ private void getStrings(HashSet<String> strHashSet, SearchCondition searchCondit
      */
     public HashSet<String> getAllStrings()
     {
-        HashSet<String> strHashSet = new LinkedHashSet<>();
+        HashSet<String> strHashSet = new LinkedHashSet<String>();
 
         if (sourceNode != null)
             getStrings(strHashSet, SearchCondition.NO_SEARCH_CONDITION, null, "", sourceNode.getOutgoingTransitions());
@@ -1066,6 +1066,6 @@ else if (nodeObjClass.equals(SimpleMDAGNode.class))
      */
     public HashMap<MDAGNode, MDAGNode> _getEquivalenceClassMDAGNodeHashMap()
     {
-        return new HashMap<>(equivalenceClassMDAGNodeHashMap);
+        return new HashMap<MDAGNode, MDAGNode>(equivalenceClassMDAGNodeHashMap);
     }
 }
diff --git a/src/main/java/com/hankcs/hanlp/collection/MDAG/MDAGMap.java b/src/main/java/com/hankcs/hanlp/collection/MDAG/MDAGMap.java
@@ -21,7 +21,7 @@
  */
 public class MDAGMap<V> extends AbstractMap<String, V>
 {
-    ArrayList<V> valueList = new ArrayList<>();
+    ArrayList<V> valueList = new ArrayList<V>();
     MDAGForMap mdag = new MDAGForMap();
 
     @Override
@@ -69,7 +69,7 @@ public Set<Entry<String, V>> entrySet()
     public Set<String> keySet()
     {
         HashSet<String> stringSet = mdag.getAllStrings();
-        LinkedHashSet<String> keySet = new LinkedHashSet<>();
+        LinkedHashSet<String> keySet = new LinkedHashSet<String>();
         Iterator<String> iterator = stringSet.iterator();
         while (iterator.hasNext())
         {
@@ -88,7 +88,7 @@ public Set<String> keySet()
     public LinkedList<Entry<String, V>> commonPrefixSearchWithValue(char[] key, int begin)
     {
         LinkedList<Entry<String, Integer>> valueIndex = mdag.commonPrefixSearchWithValueIndex(key, begin);
-        LinkedList<Entry<String, V>> entryList = new LinkedList<>();
+        LinkedList<Entry<String, V>> entryList = new LinkedList<Entry<String, V>>();
         for (Entry<String, Integer> entry : valueIndex)
         {
             entryList.add(new SimpleEntry<String, V>(entry.getKey(), valueList.get(entry.getValue())));
@@ -175,7 +175,7 @@ private int getValueIndex(MDAGNode currentNode)
 
         public LinkedList<Entry<String, Integer>> commonPrefixSearchWithValueIndex(char[] key, int begin)
         {
-            LinkedList<Map.Entry<String, Integer>> result = new LinkedList<>();
+            LinkedList<Map.Entry<String, Integer>> result = new LinkedList<Map.Entry<String, Integer>>();
             if (sourceNode != null)
             {
                 int charCount = key.length;

diff --git a/src/main/java/com/hankcs/hanlp/collection/dartsclone/DartMap.java b/src/main/java/com/hankcs/hanlp/collection/dartsclone/DartMap.java
@@ -28,7 +28,7 @@ public DartMap(List<String> keyList, List<V> valueList)
         {
             valueArray[i] = i;
         }
-        this.valueArray = new ArrayList<>(valueList);
+        this.valueArray = new ArrayList<V>(valueList);
         build(keyList, valueArray);
     }
 
@@ -85,7 +85,7 @@ public ArrayList<Pair<String, V>> commonPrefixSearch(String key, int offset, int
     {
         byte[] keyBytes = key.getBytes(utf8);
         List<Pair<Integer, Integer>> pairList = commonPrefixSearch(keyBytes, offset, maxResults);
-        ArrayList<Pair<String, V>> resultList = new ArrayList<>(pairList.size());
+        ArrayList<Pair<String, V>> resultList = new ArrayList<Pair<String, V>>(pairList.size());
         for (Pair<Integer, Integer> pair : pairList)
         {
             resultList.add(new Pair<String, V>(new String(keyBytes, 0, pair.first), valueArray.get(pair.second)));

diff --git a/src/main/java/com/hankcs/hanlp/collection/trie/bintrie/BaseNode.java b/src/main/java/com/hankcs/hanlp/collection/trie/bintrie/BaseNode.java
@@ -174,7 +174,7 @@ protected void walkToLoad(ByteArray byteArray, ValueArray valueArray)
         child = new BaseNode[childSize];
         for (int i = 0; i < childSize; ++i)
         {
-            child[i] = new Node<>();
+            child[i] = new Node<V>();
             child[i].walkToLoad(byteArray, valueArray);
         }
     }

diff --git a/src/main/java/com/hankcs/hanlp/collection/trie/bintrie/BinTrie.java b/src/main/java/com/hankcs/hanlp/collection/trie/bintrie/BinTrie.java
@@ -166,7 +166,7 @@ public Set<Map.Entry<String, V>> entrySet()
      */
     public Set<Map.Entry<String, V>> prefixSearch(String key)
     {
-        Set<Map.Entry<String, V>> entrySet = new TreeSet<>();
+        Set<Map.Entry<String, V>> entrySet = new TreeSet<Map.Entry<String, V>>();
         StringBuilder sb = new StringBuilder(key.substring(0, key.length() - 1));
         BaseNode branch = this;
         char[] chars = key.toCharArray();
@@ -202,7 +202,7 @@ public LinkedList<Map.Entry<String, V>> commonPrefixSearchWithValue(String key)
      */
     public LinkedList<Map.Entry<String, V>> commonPrefixSearchWithValue(char[] chars, int begin)
     {
-        LinkedList<Map.Entry<String, V>> result = new LinkedList<>();
+        LinkedList<Map.Entry<String, V>> result = new LinkedList<Map.Entry<String, V>>();
         StringBuilder sb = new StringBuilder();
         BaseNode branch = this;
         for (int i = begin; i < chars.length; ++i)
@@ -213,7 +213,7 @@ public LinkedList<Map.Entry<String, V>> commonPrefixSearchWithValue(char[] chars
             sb.append(aChar);
             if (branch.status == Status.WORD_MIDDLE_2 || branch.status == Status.WORD_END_3)
             {
-                result.add(new AbstractMap.SimpleEntry<>(sb.toString(), (V) branch.value));
+                result.add(new AbstractMap.SimpleEntry<String, V>(sb.toString(), (V) branch.value));
             }
         }
 
@@ -359,7 +359,7 @@ public boolean load(String path, V[] value)
             int flag = byteArray.nextInt();
             if (flag == 1)
             {
-                child[i] = new Node<>();
+                child[i] = new Node<V>();
                 child[i].walkToLoad(byteArray, valueArray);
             }
         }
@@ -385,7 +385,7 @@ public boolean load(String path)
             int flag = byteArray.nextInt();
             if (flag == 1)
             {
-                child[i] = new Node<>();
+                child[i] = new Node<V>();
                 child[i].walkToLoad(byteArray, valueArray);
             }
         }
@@ -401,7 +401,7 @@ public boolean load(ByteArray byteArray, ValueArray valueArray)
             int flag = byteArray.nextInt();
             if (flag == 1)
             {
-                child[i] = new Node<>();
+                child[i] = new Node<V>();
                 child[i].walkToLoad(byteArray, valueArray);
             }
         }

diff --git a/src/main/java/com/hankcs/hanlp/corpus/dependency/CoNll/CoNLLLoader.java b/src/main/java/com/hankcs/hanlp/corpus/dependency/CoNll/CoNLLLoader.java
@@ -23,14 +23,14 @@ public class CoNLLLoader
 {
     public static LinkedList<CoNLLSentence> loadSentenceList(String path)
     {
-        LinkedList<CoNLLSentence> result = new LinkedList<>();
-        LinkedList<CoNllLine> lineList = new LinkedList<>();
+        LinkedList<CoNLLSentence> result = new LinkedList<CoNLLSentence>();
+        LinkedList<CoNllLine> lineList = new LinkedList<CoNllLine>();
         for (String line : IOUtil.readLineListWithLessMemory(path))
         {
             if (line.trim().length() == 0)
             {
                 result.add(new CoNLLSentence(lineList));
-                lineList = new LinkedList<>();
+                lineList = new LinkedList<CoNllLine>();
                 continue;
             }
             lineList.add(new CoNllLine(line.split("\t")));

diff --git a/src/main/java/com/hankcs/hanlp/corpus/dependency/CoNll/PosTagCompiler.java b/src/main/java/com/hankcs/hanlp/corpus/dependency/CoNll/PosTagCompiler.java
@@ -27,27 +27,36 @@ public class PosTagCompiler
      */
     public static String compile(String tag, String name)
     {
-        switch (tag)
-        {
-            case "m":
-            case "mq":
-                return Predefine.TAG_NUMBER;
-            case "nr":
-            case "nr1":
-            case "nr2":
-            case "nrf":
-            case "nrj":
-                return Predefine.TAG_PEOPLE;
-            case "ns":
-            case "nsf":
-                return Predefine.TAG_PLACE;
-            case "nt":
-                return Predefine.TAG_TIME;
-            case "x":
-                return Predefine.TAG_CLUSTER;
-            case "nx":
-                return Predefine.TAG_PROPER;
-        }
+        if (tag.startsWith("m")) return Predefine.TAG_NUMBER;
+        else if (tag.startsWith("nr")) return Predefine.TAG_PEOPLE;
+        else if (tag.startsWith("ns")) return Predefine.TAG_PLACE;
+        else if (tag.startsWith("nt")) return Predefine.TAG_GROUP;
+        else if (tag.startsWith("t")) return Predefine.TAG_TIME;
+        else if (tag.equals("x")) return Predefine.TAG_CLUSTER;
+        else if (tag.equals("nx")) return Predefine.TAG_PROPER;
+        else if (tag.equals("xx")) return Predefine.TAG_OTHER;
+
+//        switch (tag)
+//        {
+//            case "m":
+//            case "mq":
+//                return Predefine.TAG_NUMBER;
+//            case "nr":
+//            case "nr1":
+//            case "nr2":
+//            case "nrf":
+//            case "nrj":
+//                return Predefine.TAG_PEOPLE;
+//            case "ns":
+//            case "nsf":
+//                return Predefine.TAG_PLACE;
+//            case "nt":
+//                return Predefine.TAG_TIME;
+//            case "x":
+//                return Predefine.TAG_CLUSTER;
+//            case "nx":
+//                return Predefine.TAG_PROPER;
+//        }
 
         return name;
     }

diff --git a/src/main/java/com/hankcs/hanlp/corpus/dependency/model/MaxEntDependencyModelMaker.java b/src/main/java/com/hankcs/hanlp/corpus/dependency/model/MaxEntDependencyModelMaker.java
@@ -43,7 +43,7 @@ public static boolean makeModel(String corpusLoadPath, String modelSavePath) thr
                 {
                     if (i == j) continue;
                     // 这就是一个边的实例，从i出发，到j，当然它可能存在也可能不存在，不存在取null照样是一个实例
-                    List<String> contextList = new LinkedList<>();
+                    List<String> contextList = new LinkedList<String>();
                     // 先生成i和j的原子特征
                     contextList.addAll(generateSingleWordContext(word, i, "i"));
                     contextList.addAll(generateSingleWordContext(word, j, "j"));
@@ -68,7 +68,7 @@ public static boolean makeModel(String corpusLoadPath, String modelSavePath) thr
 
     public static Collection<String> generateSingleWordContext(CoNLLWord[] word, int index, String mark)
     {
-        Collection<String> context = new LinkedList<>();
+        Collection<String> context = new LinkedList<String>();
         for (int i = index - 2; i < index + 2 + 1; ++i)
         {
             CoNLLWord w = i >= 0 && i < word.length ? word[i] : CoNLLWord.NULL;
@@ -81,7 +81,7 @@ public static Collection<String> generateSingleWordContext(CoNLLWord[] word, int
 
     public static Collection<String> generateUniContext(CoNLLWord[] word, int i, int j)
     {
-        Collection<String> context = new LinkedList<>();
+        Collection<String> context = new LinkedList<String>();
         context.add(word[i].NAME + '→' + word[j].NAME);
         context.add(word[i].POSTAG + '→' + word[j].POSTAG);
         context.add(word[i].NAME + '→' + word[j].NAME + (i - j));

diff --git a/src/main/java/com/hankcs/hanlp/corpus/dependency/model/WordNatureWeightModelMaker.java b/src/main/java/com/hankcs/hanlp/corpus/dependency/model/WordNatureWeightModelMaker.java
@@ -30,7 +30,7 @@ public class WordNatureWeightModelMaker
 {
     public static boolean makeModel(String corpusLoadPath, String modelSavePath)
     {
-        Set<String> posSet = new TreeSet<>();
+        Set<String> posSet = new TreeSet<String>();
         DictionaryMaker dictionaryMaker = new DictionaryMaker();
         for (CoNLLSentence sentence : CoNLLLoader.loadSentenceList(corpusLoadPath))
         {