Lucene中的Document

Posted 闫广庆

tags:

篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了Lucene中的Document相关的知识,希望对你有一定的参考价值。

所有的技术都是为了人们的应用而不是让技术变得高不可攀,Lucene也一样,Lucene是一个数据库,这个数据库中的核心就是这个document,这个东西是一种数据结构。在Lucene的document下

/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements.  See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License.  You may obtain a copy of the License at
*
*     http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.lucene.document;


import java.util.*;

import org.apache.lucene.index.IndexReader;  // for javadoc
import org.apache.lucene.index.IndexableField;
import org.apache.lucene.search.IndexSearcher;  // for javadoc
import org.apache.lucene.search.ScoreDoc; // for javadoc
import org.apache.lucene.util.BytesRef;

/** Documents are the unit of indexing and search.
*
* A Document is a set of fields.  Each field has a name and a textual value.
* A field may be {@link org.apache.lucene.index.IndexableFieldType#stored() stored} with the document, in which
* case it is returned with search hits on the document.  Thus each document
* should typically contain one or more stored fields which uniquely identify
* it.
*
* <p>Note that fields which are <i>not</i> {@link org.apache.lucene.index.IndexableFieldType#stored() stored} are
* <i>not</i> available in documents retrieved from the index, e.g. with {@link
* ScoreDoc#doc} or {@link IndexReader#document(int)}.
*/

public final class Document implements Iterable<IndexableField> {

private final List<IndexableField> fields = new ArrayList<>();

 /** Constructs a new document with no fields. */
 public Document() {}

@Override
 public Iterator<IndexableField> iterator() {
return fields.iterator();
 }

/**
  * <p>Adds a field to a document.  Several fields may be added with
  * the same name.  In this case, if the fields are indexed, their text is
  * treated as though appended for the purposes of search.</p>
  * <p> Note that add like the removeField(s) methods only makes sense
  * prior to adding a document to an index. These methods cannot
  * be used to change the content of an existing index! In order to achieve this,
  * a document has to be deleted from an index and a new changed version of that
  * document has to be added.</p>
  */
 public final void add(IndexableField field) {
fields.add(field);
 }

/**
  * <p>Removes field with the specified name from the document.
  * If multiple fields exist with this name, this method removes the first field that has been added.
  * If there is no field with the specified name, the document remains unchanged.</p>
  * <p> Note that the removeField(s) methods like the add method only make sense
  * prior to adding a document to an index. These methods cannot
  * be used to change the content of an existing index! In order to achieve this,
  * a document has to be deleted from an index and a new changed version of that
  * document has to be added.</p>
  */
 public final void removeField(String name) {
Iterator<IndexableField> it = fields.iterator();
   while (it.hasNext()) {
IndexableField field = it.next();
     if (field.name().equals(name)) {
it.remove();
       return;
     }
}
}

/**
  * <p>Removes all fields with the given name from the document.
  * If there is no field with the specified name, the document remains unchanged.</p>
  * <p> Note that the removeField(s) methods like the add method only make sense
  * prior to adding a document to an index. These methods cannot
  * be used to change the content of an existing index! In order to achieve this,
  * a document has to be deleted from an index and a new changed version of that
  * document has to be added.</p>
  */
 public final void removeFields(String name) {
Iterator<IndexableField> it = fields.iterator();
   while (it.hasNext()) {
IndexableField field = it.next();
     if (field.name().equals(name)) {
it.remove();
     }
}
}


/**
 * Returns an array of byte arrays for of the fields that have the name specified
 * as the method parameter.  This method returns an empty
 * array when there are no matching fields.  It never
 * returns null.
 *
 * @param name the name of the field
 * @return a <code>BytesRef[]</code> of binary field values
 */
 public final BytesRef[] getBinaryValues(String name) {
final List<BytesRef> result = new ArrayList<>();
   for (IndexableField field : fields) {
if (field.name().equals(name)) {
final BytesRef bytes = field.binaryValue();
       if (bytes != null) {
result.add(bytes);
       }
}
}

return result.toArray(new BytesRef[result.size()]);
 }

/**
 * Returns an array of bytes for the first (or only) field that has the name
 * specified as the method parameter. This method will return <code>null</code>
 * if no binary fields with the specified name are available.
 * There may be non-binary fields with the same name.
 *
 * @param name the name of the field.
 * @return a <code>BytesRef</code> containing the binary field value or <code>null</code>
 */
 public final BytesRef getBinaryValue(String name) {
for (IndexableField field : fields) {
if (field.name().equals(name)) {
final BytesRef bytes = field.binaryValue();
       if (bytes != null) {
return bytes;
       }
}
}
return null;
 }

/** Returns a field with the given name if any exist in this document, or
  * null.  If multiple fields exists with this name, this method returns the
  * first value added.
  */
 public final IndexableField getField(String name) {
for (IndexableField field : fields) {
if (field.name().equals(name)) {
return field;
     }
}
return null;
 }

/**
  * Returns an array of {@link IndexableField}s with the given name.
  * This method returns an empty array when there are no
  * matching fields.  It never returns null.
  *
  * @param name the name of the field
  * @return a <code>IndexableField[]</code> array
  */
 public IndexableField[] getFields(String name) {
List<IndexableField> result = new ArrayList<>();
   for (IndexableField field : fields) {
if (field.name().equals(name)) {
result.add(field);
     }
}

return result.toArray(new IndexableField[result.size()]);
 }

/** Returns a List of all the fields in a document.
  * <p>Note that fields which are <i>not</i> stored are
  * <i>not</i> available in documents retrieved from the
  * index, e.g. {@link IndexSearcher#doc(int)} or {@link
  * IndexReader#document(int)}.
  */
 public final List<IndexableField> getFields() {
return fields;
 }

private final static String[] NO_STRINGS = new String[0];

 /**
  * Returns an array of values of the field specified as the method parameter.
  * This method returns an empty array when there are no
  * matching fields.  It never returns null.
  * For {@link IntField}, {@link LongField}, {@link
  * FloatField} and {@link DoubleField} it returns the string value of the number. If you want
  * the actual numeric field instances back, use {@link #getFields}.
  * @param name the name of the field
  * @return a <code>String[]</code> of field values
  */
 public final String[] getValues(String name) {
List<String> result = new ArrayList<>();
   for (IndexableField field : fields) {
if (field.name().equals(name) && field.stringValue() != null) {
result.add(field.stringValue());
     }
}

if (result.size() == 0) {
return NO_STRINGS;
   }

return result.toArray(new String[result.size()]);
 }

/** Returns the string value of the field with the given name if any exist in
  * this document, or null.  If multiple fields exist with this name, this
  * method returns the first value added. If only binary fields with this name
  * exist, returns null.
  * For {@link IntField}, {@link LongField}, {@link
  * FloatField} and {@link DoubleField} it returns the string value of the number. If you want
  * the actual numeric field instance back, use {@link #getField}.
  */
 public final String get(String name) {
for (IndexableField field : fields) {
if (field.name().equals(name) && field.stringValue() != null) {
return field.stringValue();
     }
}
return null;
 }

/** Prints the fields of a document for human consumption. */
 @Override
 public final String toString() {
StringBuilder buffer = new StringBuilder();
   buffer.append("Document<");
   for (int i = 0; i < fields.size(); i++) {
IndexableField field = fields.get(i);
     buffer.append(field.toString());
     if (i != fields.size()-1) {
buffer.append(" ");
     }
}
buffer.append(">");
   return buffer.toString();
 }
}

我们可以看到在源码中即具有迭代器又具有两个较为特殊的类。

一本书会具有目录和正文,Lucene恰巧就是对这个结构进行的实现。


field

就是一种存储形式,在这个类里面我们将所需要索引的字段名,字段数据,是否保存进行写入。

getBinaryValues

而这个就浅显易懂的是按照我们搜索的内容从而返回我们所需要的数据集合。

BytesRef[]

这个返回的数据类型是不是有点懵,那么我们看一下相关的源码。

/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements.  See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License.  You may obtain a copy of the License at
*
*     http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.lucene.util;


import java.util.Arrays;
import java.util.Comparator;

/** Represents byte[], as a slice (offset + length) into an
*  existing byte[].  The {@link #bytes} member should never be null;
*  use {@link #EMPTY_BYTES} if necessary.
*
* <p><b>Important note:</b> Unless otherwise noted, Lucene uses this class to
* represent terms that are encoded as <b>UTF8</b> bytes in the index. To
* convert them to a Java {@link String} (which is UTF16), use {@link #utf8ToString}.
* Using code like {@code new String(bytes, offset, length)} to do this
* is <b>wrong</b>, as it does not respect the correct character set
* and may return wrong results (depending on the platform's defaults)!
*/
public final class BytesRef implements Comparable<BytesRef>,Cloneable {
/** An empty byte array for convenience */
 public static final byte[] EMPTY_BYTES = new byte[0];

 /** The contents of the BytesRef. Should never be {@code null}. */
 public byte[] bytes;

 /** Offset of first valid byte. */
 public int offset;

 /** Length of used bytes. */
 public int length;

 /** Create a BytesRef with {@link #EMPTY_BYTES} */
 public BytesRef() {
this(EMPTY_BYTES);
 }

/** This instance will directly reference bytes w/o making a copy.
  * bytes should not be null.
  */
 public BytesRef(byte[] bytes, int offset, int length) {
this.bytes = bytes;
   this.offset = offset;
   this.length = length;
   assert isValid();
 }

/** This instance will directly reference bytes w/o making a copy.
  * bytes should not be null */
 public BytesRef(byte[] bytes) {
this(bytes, 0, bytes.length);
 }

/**
  * Create a BytesRef pointing to a new array of size <code>capacity</code>.
  * Offset and length will both be zero.
  */
 public BytesRef(int capacity) {
this.bytes = new byte[capacity];
 }

/**
  * Initialize the byte[] from the UTF8 bytes
  * for the provided String.  
  *
  * @param text This must be well-formed
  * unicode text, with no unpaired surrogates.
  */
 public BytesRef(CharSequence text) {
this(new byte[UnicodeUtil.MAX_UTF8_BYTES_PER_CHAR * text.length()]);
   length = UnicodeUtil.UTF16toUTF8(text, 0, text.length(), bytes);
 }

/**
  * Expert: compares the bytes against another BytesRef,
  * returning true if the bytes are equal.
  *
  * @param other Another BytesRef, should not be null.
  * @lucene.internal
  */
 public boolean bytesEquals(BytesRef other) {
assert other != null;
   if (length == other.length) {
int otherUpto = other.offset;
     final byte[] otherBytes = other.bytes;
     final int end = offset + length;
     for(int upto=offset;upto<end;upto++,otherUpto++) {
if (bytes[upto] != otherBytes[otherUpto]) {
return false;
       }
}
return true;
   } else {
return false;
   }
}

/**
  * Returns a shallow clone of this instance (the underlying bytes are
  * <b>not</b> copied and will be shared by both the returned object and this
  * object.
  *
  * @see #deepCopyOf
  */
 @Override
 public BytesRef clone() {
return new BytesRef(bytes, offset, length);
 }

/** Calculates the hash code as required by TermsHash during indexing.
  *  <p> This is currently implemented as MurmurHash3 (32
  *  bit), using the seed from {@link
  *  StringHelper#GOOD_FAST_HASH_SEED}, but is subject to
  *  change from release to release. */
 @Override
 public int hashCode() {
return StringHelper.murmurhash3_x86_32(this, StringHelper.GOOD_FAST_HASH_SEED);
 }

@Override
 public boolean equals(Object other) {
if (other == null) {
return false;
   }
if (other instanceof BytesRef) {
return this.bytesEquals((BytesRef) other);
   }
return false;
 }

/** Interprets stored bytes as UTF8 bytes, returning the
  *  resulting string */
 public String utf8ToString() {
final char[] ref = new char[length];
   final int len = UnicodeUtil.UTF8toUTF16(bytes, offset, length, ref);
   return new String(ref, 0, len);
 }

/** Returns hex encoded bytes, eg [0x6c 0x75 0x63 0x65 0x6e 0x65] */
 @Override
 public String toString() {
StringBuilder sb = new StringBuilder();
   sb.append('[');
   final int end = offset + length;
   for(int i=offset;i<end;i++) {
if (i > offset) {
sb.append(' ');
     }
sb.append(Integer.toHexString(bytes[i]&0xff));
   }
sb.append(']');
   return sb.toString();
 }

/** Unsigned byte order comparison */
 @Override
 public int compareTo(BytesRef other) {
return utf8SortedAsUnicodeSortOrder.compare(this, other);
 }

private final static Comparator<BytesRef> utf8SortedAsUnicodeSortOrder = new UTF8SortedAsUnicodeComparator();

 public static Comparator<BytesRef> getUTF8SortedAsUnicodeComparator() {
return utf8SortedAsUnicodeSortOrder;
 }

private static class UTF8SortedAsUnicodeComparator implements Comparator<BytesRef> {
// Only singleton
   private UTF8SortedAsUnicodeComparator() {};

   @Override
   public int compare(BytesRef a, BytesRef b) {
final byte[] aBytes = a.bytes;
     int aUpto = a.offset;
     final byte[] bBytes = b.bytes;
     int bUpto = b.offset;
     
     final int aStop = aUpto + Math.min(a.length, b.length);
     while(aUpto < aStop) {
int aByte = aBytes[aUpto++] & 0xff;
       int bByte = bBytes[bUpto++] & 0xff;

       int diff = aByte - bByte;
       if (diff != 0) {
return diff;
       }
}

// One is a prefix of the other, or, they are equal:
     return a.length - b.length;
   }
}

/** @deprecated This comparator is only a transition mechanism */
 @Deprecated
 private final static Comparator<BytesRef> utf8SortedAsUTF16SortOrder = new UTF8SortedAsUTF16Comparator();

 /** @deprecated This comparator is only a transition mechanism */
 @Deprecated
 public static Comparator<BytesRef> getUTF8SortedAsUTF16Comparator() {
return utf8SortedAsUTF16SortOrder;
 }

/** @deprecated This comparator is only a transition mechanism */
 @Deprecated
 private static class UTF8SortedAsUTF16Comparator implements Comparator<BytesRef> {
// Only singleton
   private UTF8SortedAsUTF16Comparator() {};

   @Override
   public int compare(BytesRef a, BytesRef b) {

final byte[] aBytes = a.bytes;
     int aUpto = a.offset;
     final byte[] bBytes = b.bytes;
     int bUpto = b.offset;
     
     final int aStop;
     if (a.length < b.length) {
aStop = aUpto + a.length;
     } else {
aStop = aUpto + b.length;
     }

while(aUpto < aStop) {
int aByte = aBytes[aUpto++] & 0xff;
       int bByte = bBytes[bUpto++] & 0xff;

       if (aByte != bByte) {

// See http://icu-project.org/docs/papers/utf16_code_point_order.html#utf-8-in-utf-16-order

         // We know the terms are not equal, but, we may
         // have to carefully fixup the bytes at the
         // difference to match UTF16's sort order:
         
         // NOTE: instead of moving supplementary code points (0xee and 0xef) to the unused 0xfe and 0xff,
         // we move them to the unused 0xfc and 0xfd [reserved for future 6-byte character sequences]
         // this reserves 0xff for preflex's term reordering (surrogate dance), and if unicode grows such
         // that 6-byte sequences are needed we have much bigger problems anyway.
         if (aByte >= 0xee && bByte >= 0xee) {
if ((aByte & 0xfe) == 0xee) {
aByte += 0xe;
           }
if ((bByte&0xfe) == 0xee) {
bByte += 0xe;
           }
}
return aByte - bByte;
       }
}

// One is a prefix of the other, or, they are equal:
     return a.length - b.length;
   }
}

/**
  * Creates a new BytesRef that points to a copy of the bytes from
  * <code>other</code>
  * <p>
  * The returned BytesRef will have a length of other.length
  * and an offset of zero.
  */
 public static BytesRef deepCopyOf(BytesRef other) {
BytesRef copy = new BytesRef();
   copy.bytes = Arrays.copyOfRange(other.bytes, other.offset, other.offset + other.length);
   copy.offset = 0;
   copy.length = other.length;
   return copy;
 }

/**
  * Performs internal consistency checks.
  * Always returns true (or throws IllegalStateException)
  */
 public boolean isValid() {
if (bytes == null) {
throw new IllegalStateException("bytes is null");
   }
if (length < 0) {
throw new IllegalStateException("length is negative: " + length);
   }
if (length > bytes.length) {
throw new IllegalStateException("length is out of bounds: " + length + ",bytes.length=" + bytes.length);
   }
if (offset < 0) {
throw new IllegalStateException("offset is negative: " + offset);
   }
if (offset > bytes.length) {
throw new IllegalStateException("offset out of bounds: " + offset + ",bytes.length=" + bytes.length);
   }
if (offset + length < 0) {
throw new IllegalStateException("offset+length is negative: offset=" + offset + ",length=" + length);
   }
if (offset + length > bytes.length) {
throw new IllegalStateException("offset+length out of bounds: offset=" + offset + ",length=" + length + ",bytes.length=" + bytes.length);
   }
return true;
 }
}

这里面我们能看到这就是对我们数据库中的字节流进行编解码的一个工具,我现在其实更多的就是因为卡在这个环节无法自拔。

以上是关于Lucene中的Document的主要内容,如果未能解决你的问题,请参考以下文章

Lucene的QueryParser

Lucene(7.3.1)学习笔记-Document类源码解析

搜索引擎系列五:Lucene索引详解(IndexWriter详解Document详解索引更新)

lucene中创建索引库

Lucene全文检索

Lucene Document和Java ArrayList