solrj:org.apache.solr.common.util.NamedList.java

package org.apache.solr.common.util;
 
import java.io.Serializable;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
import java.util.Iterator;
import java.util.List;
import java.util.Map;
 
import org.apache.solr.common.SolrException;
 
/**
 * 一个简单的容器类  用来模拟一个有序的 name/value 键值对列表
 * <p>
 *  不同于 Maps:
 * </p>
 * <ul>
 *  <li>Names可以重复</li>
 *  <li>元素具有顺序性</li>
 *  <li>元素可以通过数值索引来访问</li>
 *  <li>Names和 Values可以都为 null</li>
 * </ul>
 *
 * <p>
 * A NamedList provides fast access by element number, but not by name.
 * </p>
 * <p>
 * NamedList被序列化后,元素的顺序就比较重要了.所以ResponseWriters输出格式如JSON通常会选择一个容易在不同客户端保存并持有顺序的数据结构.
 *
 * 如果通过key访问比序列化更重要是,参考{@link SimpleOrderedMap},或者简单使用常规的{@link Map}
 * </p>
 *
 */
public class NamedList<T> implements Cloneable, Serializable, Iterable<Map.Entry<String,T>> {
 
  private static final long serialVersionUID = 1957981902839867821L;
  protected final List<Object> nvPairs;
 
  /** Creates an empty instance */
  public NamedList() {
    nvPairs = new ArrayList<Object>();
  }
 
  /**
   *创建一个实例,支持Map.Entry<String, ? extends T>[]类型
   *
   * <p>
   * Modifying the contents of the Entry[] after calling this constructor may change
   * the NamedList (in future versions of Solr), but this is not guaranteed and should
   * not be relied upon.  To modify the NamedList, refer to {@link #add(String, Object)}
   * or {@link #remove(String)}.
   * </p>
   *
   * @param nameValuePairs the name value pairs
   */
  public NamedList(Map.Entry<String, ? extends T>[] nameValuePairs) {
    nvPairs = nameValueMapToList(nameValuePairs);
  }
 
  /**
   *创建一个实例,支持明确的name/value配对键值.
   * <p>
   * When using this constructor, runtime type safety is only guaranteed if
   * all even numbered elements of the input list are of type "T".
   * </p>
   *
   * @param nameValuePairs underlying List which should be used to implement a NamedList
   * @deprecated Use {@link #NamedList(java.util.Map.Entry[])} for the NamedList instantiation
   */
  @Deprecated
  public NamedList(List<Object> nameValuePairs) {
    nvPairs=nameValuePairs;
  }
 
  /**
   *
   * 序列化Map.Entry&lt;String,?&gt; 为一个List.这个List中索引为(0,2,4. ..etc)的是String,奇数元素(1,3,5...etc)为"T"类型.
   *
   * @return Modified List as per the above description
   * @deprecated This a temporary placeholder method until the guts of the class
   * are actually replaced by List&lt;String, ?&gt;.
   * @see <a href="https://issues.apache.org/jira/browse/SOLR-912">SOLR-912</a>
   */
  @Deprecated
  private List<Object> nameValueMapToList(Map.Entry<String, ? extends T>[] nameValuePairs) {
    List<Object> result = new ArrayList<Object>();
    for (Map.Entry<String, ?> ent : nameValuePairs) {
      result.add(ent.getKey());
      result.add(ent.getValue());
    }
    return result;
  }
 
  /** The total number of name/value pairs */
  public int size() {
    return nvPairs.size() >> 1;
  }
 
  /**
   * The name of the pair at the specified List index
   *
   * @return null if no name exists
   */
  public String getName(int idx) {
    return (String)nvPairs.get(idx << 1);
  }
 
  /**
   * The value of the pair at the specified List index
   *
   * @return may be null
   */
  @SuppressWarnings("unchecked")
  public T getVal(int idx) {
    return (T)nvPairs.get((idx << 1) + 1);
  }
 
  /**
   * 在list的末端添加 name/value键值对.
   */
  public void add(String name, T val) {
    nvPairs.add(name);
    nvPairs.add(val);
  }
 
  /**
   * 修改指定索引处的键值对的name值.
   */
  public void setName(int idx, String name) {
    nvPairs.set(idx<<1, name);
  }
 
  /**
   *修改指定索引处的键值对的value值.
   *
   * @return 老的对于索引的value值
   */
  public T setVal(int idx, T val) {
    int index = (idx<<1)+1;
    @SuppressWarnings("unchecked")
    T old = (T)nvPairs.get( index );
    nvPairs.set(index, val);
    return old;
  }
 
  /**
   *删除指定索引处的键值对的name/value值.
   *
   * @return 删除的键值对的value值
   */
  public T remove(int idx) {
    int index = (idx<<1);
    nvPairs.remove(index);
    @SuppressWarnings("unchecked")
    T result = (T)nvPairs.remove(index);  // same index, as things shifted in previous remove
    return result;
  }
 
  /**
   * 扫描指定索引处开始的List列表,并返回第一处name为指定名字的键值对的索引
   *
   * @param name 查询的name,可能为null
   * @param start 搜索查询起始索引
   * @return 第一处匹配键值的索引,如果不匹配,返回-1
   */
  public int indexOf(String name, int start) {
    int sz = size();
    for (int i=start; i<sz; i++) {
      String n = getName(i);
      if (name==null) {
        if (n==null) return i; // matched null
      } else if (name.equals(n)) {
        return i;
      }
    }
    return -1;
  }
 
  /**
   * 返回第一个name为指定值的实例的value值.
   *
   * <p>
   * NOTE: this runs in linear time (it scans starting at the
   * beginning of the list until it finds the first pair with
   * the specified name).
   *
   * @return null if not found or if the value stored was null.
   * @see #indexOf
   * @see #get(String,int)
   *
   */
  public T get(String name) {
    return get(name,0);
  }
 
  /**
   * Gets the value for the first instance of the specified name
   * found starting at the specified index.
   * <p>
   * NOTE: this runs in linear time (it scans starting at the
   * specified position until it finds the first pair with
   * the specified name).
   *
   * @return null if not found or if the value stored was null.
   * @see #indexOf
   */
  public T get(String name, int start) {
    int sz = size();
    for (int i=start; i<sz; i++) {
      String n = getName(i);
      if (name==null) {
        if (n==null) return getVal(i);
      } else if (name.equals(n)) {
        return getVal(i);
      }
    }
    return null;
  }
 
  /**
   * Gets the values for the the specified name
   *
   * @param name Name
   * @return List of values
   */
  public List<T> getAll(String name) {
    List<T> result = new ArrayList<T>();
    int sz = size();
    for (int i = 0; i < sz; i++) {
      String n = getName(i);
      if (name==n || (name!=null && name.equals(n))) {
        result.add(getVal(i));
      }
    }
    return result;
  }
 
  /**
   * Removes all values matching the specified name
   *
   * @param name Name
   */
  private void killAll(String name) {
    int sz = size();
    // Go through the list backwards, removing matches as found.
    for (int i = sz - 1; i >= 0; i--) {
      String n = getName(i);
      if (name==n || (name!=null && name.equals(n))) {
        remove(i);
      }
    }
  }
 
  /**
   * 递归解析NameList结构到一个指定的元素中.随着NameList树的解析,最后一个元素可以是任何类型,包括NameList,
   * 但前面所有元素必须NamedList对象本身.如果指定的层次结构不存在,那么返回为null.NameList是允许null值的,
   * 所以最后返回的值也可以是null;
   *
   * 这个方法对解析solr响应的/admin/mbeans 句柄特别有用,当然也同样用于其他复杂结构的工作处理.
   *
   * 推荐明确抛出返回值,一个比较安全的选择及时接受Object对象的返回值,然后去确认它的类型.
   *
   * 使用示例:
   *
   * String coreName = (String) response.findRecursive
   * ("solr-mbeans", "CORE", "core", "stats", "coreName");
   * long numDoc = (long) response.findRecursive
   * ("solr-mbeans", "CORE", "searcher", "stats", "numDocs");
   *
   * @param args
   *          One or more strings specifying the tree to navigate.
   * @return the last entry in the given path hierarchy, null if not found.
   */
  public Object findRecursive(String... args) {
    NamedList<?> currentList = null;
    Object value = null;
    for (int i = 0; i < args.length; i++) {
      String key = args[i];
      /*
       * 第一次循环,currentList为null,所以我们把 NameList这个对象分配给currentList.
       * 然后我们检索这个列表的第一个key,然后把key对应的对象值 赋值给value 变量.
       *
       * 第二次循环遍历时,首先确认上一次我们获得的value是否是一个NameList.
       * 如果是NameList对象,那么将该对象赋值给currentList,抓取下一个key对应的value值,并开始遍历.
       * 如果不是一个NameList对象,重置value值为null,中断遍历.
       *
       * 赋值value为null,然后结束循环遍历,看起来做的不正确,但是有一个非常简单的原因,
       * 它的工作原理:如果循环到最后一个key,在检索到对应的value值时会自然结束循环遍历,并且这段代码永远不会执行的.
       *
       */
      if (currentList == null) {
        currentList = this;
      } else {
        if (value instanceof NamedList) {
          currentList = (NamedList<?>) value;
        } else {
          value = null;
          break;
        }
      }
      /*
       * 这里不再需要验证currentList是否为null.如果当前list为null的话,上面代码value instanceof NamedList
       * 的检查会失败的.如果这种情况发生的话,循环就会在这之前结束.
       *
       */
      value = currentList.get(key, 0);
    }
    return value;
  }
 
  @Override
  public String toString() {
    StringBuilder sb = new StringBuilder();
    sb.append('{');
    int sz = size();
    for (int i=0; i<sz; i++) {
      if (i != 0) sb.append(',');
      sb.append(getName(i));
      sb.append('=');
      sb.append(getVal(i));
    }
    sb.append('}');
 
    return sb.toString();
  }
 
  /**
   *
   * 帮助类实现 Map.Entry&lt;String, T&gt;用来存储 NamedList中的key-value关系.
   */
  public static final class NamedListEntry<T> implements Map.Entry<String,T> {
 
    public NamedListEntry() {
 
    }
 
    public NamedListEntry(String _key, T _value) {
      key = _key;
      value = _value;
    }
 
    @Override
    public String getKey() {
      return key;
    }
 
    @Override
    public T getValue() {
      return value;
    }
 
    @Override
    public T setValue(T _value) {
      T oldValue = value;
      value = _value;
      return oldValue;
    }
 
    private String key;
 
    private T value;
  }
 
  /**
   * 遍历Map，依次增加了它的键/值对
   *
   */
  public boolean addAll(Map<String,T> args) {
    for (Map.Entry<String, T> entry : args.entrySet() ) {
      add(entry.getKey(), entry.getValue());
    }
    return args.size()>0;
  }
 
  /**将给定的NameList的元素添加到当前NameList对象中 */
  public boolean addAll(NamedList<T> nl) {
    nvPairs.addAll(nl.nvPairs);
    return nl.size()>0;
  }
 
  /**
   * 生成一个<i>浅拷贝</i>的NameList.
   */
  @Override
  public NamedList<T> clone() {
    ArrayList<Object> newList = new ArrayList<Object>(nvPairs.size());
    newList.addAll(nvPairs);
    return new NamedList<T>(newList);
  }
 
  //----------------------------------------------------------------------------
  // Iterable 接口
  //----------------------------------------------------------------------------
 
  /**
   * 支持Iterable接口
   */
  @Override
  public Iterator<Map.Entry<String,T>> iterator() {
 
    final NamedList<T> list = this;
 
    Iterator<Map.Entry<String,T>> iter = new Iterator<Map.Entry<String,T>>() {
 
      int idx = 0;
 
      @Override
      public boolean hasNext() {
        return idx < list.size();
      }
 
      @Override
      public Map.Entry<String,T> next() {
        final int index = idx++;
        Map.Entry<String,T> nv = new Map.Entry<String,T>() {
          @Override
          public String getKey() {
            return list.getName( index );
          }
 
          @Override
          public T getValue() {
            return list.getVal( index );
          }
 
          @Override
          public String toString() {
            return getKey()+"="+getValue();
          }
 
          @Override
          public T setValue(T value) {
            return list.setVal(index, value);
          }
        };
        return nv;
      }
 
      @Override
      public void remove() {
        throw new UnsupportedOperationException();
      }
    };
    return iter;
  }
 
  /**
   * NOTE: 线性时间执行 (it scans starting at the
   * beginning of the list until it finds the first pair with
   * the specified name).
   */
  public T remove(String name) {
    int idx = indexOf(name, 0);
    if(idx != -1) return remove(idx);
    return null;
  }
 
  /**
   * 删除并返回指定name的所有values.如果没有匹配,返回null.这个方法返回所有匹配对象,不考虑数据类型.
   * 如果解析solr配置选项,{@link #removeConfigArgs(String)} 或者 {@link #removeBooleanArg(String)}
   * 方法是一个更好的选择.
   *
   * @param name Name
   * @return List of values
   */
  public List<T> removeAll(String name) {
    List<T> result = new ArrayList<T>();
    result = getAll(name);
    if (result.size() > 0 ) {
      killAll(name);
      return result;
    }
    return null;
  }
 
  /**
   * 用来从NameList对象中获取一个boolean参数.如果name不存在,返回null.如果对应name有多个value,
   * 或者这个值不是Boolean或者String类型,会抛出一个异常.如果只有一个值存在,并且是 一个Boolean或者String类型
   * 这个value值就会删除并返回一个Boolean值.如果抛出异常,NamedList将不会改变.
   * 参考 {@link #removeAll(String)}和 {@link #removeConfigArgs(String)}的更多方式:从NamedList收集的配置信息
   *
   * @param name NameList中要查询的key值.
   *
   * @return The boolean value found.
   * @throws SolrException
   *           If multiple values are found for the name or the value found is
   *           not a Boolean or a String.
   */
  public Boolean removeBooleanArg(final String name) {
    Boolean bool;
    List<T> values = getAll(name);
    if (0 == values.size()) {
      return null;
    }
    if (values.size() > 1) {
      throw new SolrException(SolrException.ErrorCode.SERVER_ERROR,
          "Only one '" + name + "' is allowed");
    }
    Object o = get(name);
    if (o instanceof Boolean) {
      bool = (Boolean)o;
    } else if (o instanceof CharSequence) {
      bool = Boolean.parseBoolean(o.toString());
    } else {
      throw new SolrException(SolrException.ErrorCode.SERVER_ERROR,
          "'" + name + "' must have type 'bool' or 'str'; found " + o.getClass());
    }
    remove(name);
    return bool;
  }
 
  /**
   * 从NamedList对象中获取一个或者多个参数用来保存配置参数.找到所有匹配给定name的entries.
   * 如果都是字符串或者字符串数组,从NameList中删除它们,并返回{@link Collection}.
   *
   * 如果返回的集合是一个{@link ArrayList},那么参数顺序将被保存.如果关联name的value值不是字符串或者是字符串数组,
   * 那么抛出一个SolrException异常.异常抛出,NameList不会更改.如果没有匹配的值,返回一个空的集合.如果需要删除,
   * 并在检索到所有匹配的条目时不考虑数据类型,那么使用 {@link #removeAll(String)} 替代.
   * {@link #removeBooleanArg(String)} 方法用来检索一个boolean参数.
   *
   * @param name  NameList中要查询的key值
   * @return A collection of the values found.
   * @throws SolrException
   *           If values are found for the input key that are not strings or
   *           arrays of strings.
   */
  @SuppressWarnings("rawtypes")
  public Collection<String> removeConfigArgs(final String name)
      throws SolrException {
    List<T> objects = getAll(name);
    List<String> collection = new ArrayList<String>(size() / 2);
    final String err = "init arg '" + name + "' must be a string "
        + "(ie: 'str'), or an array (ie: 'arr') containing strings; found: ";
 
    for (Object o : objects) {
      if (o instanceof String) {
        collection.add((String) o);
        continue;
      }
 
      // If it's an array, convert to List (which is a Collection).
      if (o instanceof Object[]) {
        o = Arrays.asList((Object[]) o);
      }
 
      // If it's a Collection, collect each value.
      if (o instanceof Collection) {
        for (Object item : (Collection) o) {
          if (!(item instanceof String)) {
            throw new SolrException(SolrException.ErrorCode.SERVER_ERROR, err + item.getClass());
          }
          collection.add((String) item);
        }
        continue;
      }
      throw new SolrException(SolrException.ErrorCode.SERVER_ERROR, err + o.getClass());
    }
 
    if (collection.size() > 0) {
      killAll(name);
    }
 
    return collection;
  }
 
  public void clear() {
    nvPairs.clear();
  }
 
  @Override
  public int hashCode() {
    return nvPairs.hashCode();
  }
 
  @Override
  public boolean equals(Object obj) {
    if (!(obj instanceof NamedList)) return false;
    NamedList<?> nl = (NamedList<?>) obj;
    return this.nvPairs.equals(nl.nvPairs);
  }
}

测试类:

package org.apache.solr.common.util;
 
import java.util.ArrayList;
import java.util.List;
 
import org.apache.lucene.util.LuceneTestCase;
import org.apache.solr.common.SolrException;
 
public class NamedListTest extends LuceneTestCase {
  public void testRemove() {
    NamedList<String> nl = new NamedList<String>();
    nl.add("key1", "value1");
    nl.add("key2", "value2");
    assertEquals(2, nl.size());
    String value = null;
    value = nl.remove(0);
    assertEquals("value1", value);
    assertEquals(1, nl.size());
    value = nl.remove("key2");
    assertEquals("value2", value);
    assertEquals(0, nl.size());
  }
 
  public void testRemoveAll() {
    NamedList<String> nl = new NamedList<String>();
    nl.add("key1", "value1-1");
    nl.add("key2", "value2-1");
    nl.add("key1", "value1-2");
    nl.add("key2", "value2-2");
    nl.add("key1", "value1-3");
    nl.add("key2", "value2-3");
    nl.add("key1", "value1-4");
    nl.add("key2", "value2-4");
    nl.add("key1", "value1-5");
    nl.add("key2", "value2-5");
    nl.add("key1", "value1-6");
    assertEquals(11, nl.size());
    List<String> values = null;
    values = nl.removeAll("key1");
    assertEquals("value1-1", values.get(0));
    assertEquals("value1-3", values.get(2));
    assertEquals(6, values.size());
    assertEquals(5, nl.size());
    values = nl.removeAll("key2");
    assertEquals(5, values.size());
    assertEquals(0, nl.size());
  }
 
  public void testRemoveArgs() {
    NamedList<Object> nl = new NamedList<Object>();
    nl.add("key1", "value1-1");
    nl.add("key2", "value2-1");
    nl.add("key1", "value1-2");
    nl.add("key2", "value2-2");
    nl.add("key1", "value1-3");
    nl.add("key2", "value2-3");
    nl.add("key1", "value1-4");
    nl.add("key2", "value2-4");
    nl.add("key1", "value1-5");
    nl.add("key2", "value2-5");
    nl.add("key1", "value1-6");
    nl.add("key2", 0);
    nl.add("key2", "value2-7");
    assertEquals(13, nl.size());
    List<String> values = (ArrayList<String>) nl.removeConfigArgs("key1");
    assertEquals("value1-1", values.get(0));
    assertEquals("value1-3", values.get(2));
    assertEquals(6, values.size());
    assertEquals(7, nl.size());
    try {
      values = (ArrayList<String>) nl.removeConfigArgs("key2");
      fail();
    }
    catch(SolrException e) {
      // Expected exception.
      assertTrue(true);
    }
    // nl should be unmodified when removeArgs throws an exception.
    assertEquals(7, nl.size());
  }
 
  public void testRecursive() {
      /**
       * NL结构说明
       */
        // key1
        // key2
        // - key2a
        // - key2b
        // --- key2b1
        // --- key2b2
        // - key2c
        // - k2int1
        // key3
        // - key3a
        // --- key3a1
        // --- key3a2
        // --- key3a3
        // - key3b
        // - key3c
 
    // 实例化一个多样的NL结构.
    NamedList<String> nl2b = new NamedList<String>();
    nl2b.add("key2b1", "value2b1");
    nl2b.add("key2b2", "value2b2");
    NamedList<String> nl3a = new NamedList<String>();
    nl3a.add("key3a1", "value3a1");
    nl3a.add("key3a2", "value3a2");
    nl3a.add("key3a3", "value3a3");
    NamedList<Object> nl2 = new NamedList<Object>();
    nl2.add("key2a", "value2a");
    nl2.add("key2b", nl2b);
    nl2.add("k2int1", (int) 5);
    NamedList<Object> nl3 = new NamedList<Object>();
    nl3.add("key3a", nl3a);
    nl3.add("key3b", "value3b");
    nl3.add("key3c", "value3c");
    nl3.add("key3c", "value3c2");
    NamedList<Object> nl = new NamedList<Object>();
    nl.add("key1", "value1");
    nl.add("key2", nl2);
    nl.add("key3", nl3);
 
    // 简单的三级检查.
    String test1 = (String) nl.findRecursive("key2", "key2b", "key2b2");
    assertEquals("value2b2", test1);
    String test2 = (String) nl.findRecursive("key3", "key3a", "key3a3");
    assertEquals("value3a3", test2);
    // 二级检查
    String test3 = (String) nl.findRecursive("key3", "key3c");
    assertEquals("value3c", test3);
    // 检查无效值返回null.
    String test4 = (String) nl.findRecursive("key3", "key3c", "invalid");
    assertEquals(null, test4);
    String test5 = (String) nl.findRecursive("key3", "invalid", "invalid");
    assertEquals(null, test5);
    String test6 = (String) nl.findRecursive("invalid", "key3c");
    assertEquals(null, test6);
    // 验证检索NamedList对象具有正确的类型.
    Object test7 = nl.findRecursive("key2", "key2b");
    assertTrue(test7 instanceof NamedList);
    // Integer检查.
    int test8 = (Integer) nl.findRecursive("key2", "k2int1");
    assertEquals(5, test8);
    // Check that a single argument works the same as get(String).
    String test9 = (String) nl.findRecursive("key1");
    assertEquals("value1", test9);
    // enl == 明确嵌套列表
    //
    // key1
    // - key1a
    // - key1b
    // key2 (null list)
    NamedList<NamedList<String>> enl = new NamedList<NamedList<String>>();
    NamedList<String> enlkey1 = new NamedList<String>();
    NamedList<String> enlkey2 = null;
    enlkey1.add("key1a", "value1a");
    enlkey1.add("key1b", "value1b");
    enl.add("key1", enlkey1);
    enl.add("key2", enlkey2);
 
    // 和上面的测试很类似, 只是重复了明确嵌套的对象类型.
    String enltest1 = (String) enl.findRecursive("key1", "key1a");
    assertEquals("value1a", enltest1);
    String enltest2 = (String) enl.findRecursive("key1", "key1b");
    assertEquals("value1b", enltest2);
    // 验证:在存储一个null值时,get方法返回的是一个null,那么验证这个递归方法.
    Object enltest3 = enl.get("key2");
    assertNull(enltest3);
    Object enltest4 = enl.findRecursive("key2");
    assertNull(enltest4);
  }
}