Apache Tika源码研究（六）

色破飞机

　　上文还没有来得及分析Apache Tika是怎样检测文档的mime类型的，以及怎样根据mime类型找到相应的Parser解析类的，下面接着说
　　在tika-parsers.jar路径文件META-INF/services/org.apache.tika.detect.Detector记录了tika提供的mime类型检测类，当然tika还有部分mime类型检测类该文件并没有记录，后面我通过分析源码可以获知。
　　该文件包含的检测类我们先睹为快：



#  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.
org.apache.tika.parser.microsoft.POIFSContainerDetector
org.apache.tika.parser.pkg.ZipContainerDetector
　　注意还有vorbis-java-tika-X.jar的同名路径下也存在该文件，tika都会加载进来，所以共包含了三个实现类
　　org.apache.tika.parser.microsoft.POIFSContainerDetector
　　org.apache.tika.parser.pkg.ZipContainerDetector
　　org.gagravarr.tika.OggDetector
　　这些tika文档mime类型检测类共同实现了Detector接口：
　　最重要的文件的mime类型检测相关接口和类的UML图如下：

　　Detector接口源码：



/**
* Content type detector. Implementations of this interface use various
* heuristics to detect the content type of a document based on given
* input metadata or the first few bytes of the document stream.
*
* @since Apache Tika 0.3
*/
public interface Detector extends Serializable {
/**
* Detects the content type of the given input document. Returns
* application/octet-stream if the type of the document
* can not be detected.
*
* If the document input stream is not available, then the first
* argument may be null. Otherwise the detector may
* read bytes from the start of the stream to help in type detection.
* The given stream is guaranteed to support the
* {@link InputStream#markSupported() mark feature} and the detector
* is expected to {@link InputStream#mark(int) mark} the stream before
* reading any bytes from it, and to {@link InputStream#reset() reset}
* the stream before returning. The stream must not be closed by the
* detector.
*
* The given input metadata is only read, not modified, by the detector.
*
* @param input document input stream, or null
* @param metadata input metadata for the document
* @return detected media type, or application/octet-stream
* @throws IOException if the document input stream could not be read
*/
MediaType detect(InputStream input, Metadata metadata) throws IOException;
}
　　
　　实现该接口的最重要的类是CompositeDetector，该类并不提供具体的mime类型检测，而是调用其他的实现类进行mime类型检测，供tika其他类调用



/**
* Content type detector that combines multiple different detection mechanisms.
*/
public class CompositeDetector implements Detector {
/**
* Serial version UID
*/
private static final long serialVersionUID = 5980683158436430252L;
private final MediaTypeRegistry registry;
private final List detectors;
public CompositeDetector(
MediaTypeRegistry registry, List detectors) {
this.registry = registry;
this.detectors = detectors;
}
public CompositeDetector(List detectors) {
this(new MediaTypeRegistry(), detectors);
}
public CompositeDetector(Detector... detectors) {
this(Arrays.asList(detectors));
}
public MediaType detect(InputStream input, Metadata metadata)
throws IOException {
MediaType type = MediaType.OCTET_STREAM;
for (Detector detector : getDetectors()) {
MediaType detected = detector.detect(input, metadata);
if (registry.isSpecializationOf(detected, type)) {
type = detected;
}
}
return type;
}
/**
* Returns the component detectors.
*/
public List getDetectors() {
return Collections.unmodifiableList(detectors);
}
}
　　
　　构造函数CompositeDetector(MediaTypeRegistry registry, List detectors)用于初始化成员变量MediaTypeRegistry registry和List detectors
　　MediaTypeRegistry registry成员注册了系统提供的mime类型，List detectors成员为系统的Detector实现类集合
　　MediaType detect(InputStream input, Metadata metadata)方法遍历Detector集合检测InputStream input的mime类型
　　CompositeDetector还有一个派生类DefaultDetector，用于初始化CompositeDetector的成员变量



public class DefaultDetector extends CompositeDetector {
/** Serial version UID */
private static final long serialVersionUID = -8170114575326908027L;
/**
* Finds all statically loadable detectors and sort the list by name,
* rather than discovery order. Detectors are used in the given order,
* so put the Tika parsers last so that non-Tika (user supplied)
* parsers can take precedence.
*
* @param loader service loader
* @return ordered list of statically loadable detectors
*/
private static List getDefaultDetectors(
MimeTypes types, ServiceLoader loader) {
List detectors =
loader.loadStaticServiceProviders(Detector.class);
Collections.sort(detectors, new Comparator() {
public int compare(Detector d1, Detector d2) {
String n1 = d1.getClass().getName();
String n2 = d2.getClass().getName();
boolean t1 = n1.startsWith("org.apache.tika.");
boolean t2 = n2.startsWith("org.apache.tika.");
if (t1 == t2) {
return n1.compareTo(n2);
} else if (t1) {
return 1;
} else {
return -1;
}
}
});
// Finally the Tika MimeTypes as a fallback
        detectors.add(types);
return detectors;
}
private transient final ServiceLoader loader;
public DefaultDetector(MimeTypes types, ServiceLoader loader) {
super(types.getMediaTypeRegistry(), getDefaultDetectors(types, loader));
this.loader = loader;
}
public DefaultDetector(MimeTypes types, ClassLoader loader) {
this(types, new ServiceLoader(loader));
}
public DefaultDetector(ClassLoader loader) {
this(MimeTypes.getDefaultMimeTypes(), loader);
}
public DefaultDetector(MimeTypes types) {
this(types, new ServiceLoader());
}
public DefaultDetector() {
this(MimeTypes.getDefaultMimeTypes());
}
@Override
public List getDetectors() {
if (loader != null) {
List detectors =
loader.loadDynamicServiceProviders(Detector.class);
detectors.addAll(super.getDetectors());
return detectors;
} else {
return super.getDetectors();
}
}
}
　　

List getDefaultDetectors(MimeTypes types, ServiceLoader loader)方法加载静态的Detector实现类，而List getDetectors()方法加载动态的Detector实现类并包含父类的Detector实现类集合
我们这里注意到，前者额外调用了detectors.add(types)，将MimeTypes types对象也添加到集合里面，因为MimeTypes类是实现了Detector接口的，前面文章我已经提到过。
所以实际用到的解析类包括四个
　　org.apache.tika.parser.microsoft.POIFSContainerDetector
　　org.apache.tika.parser.pkg.ZipContainerDetector
　　org.gagravarr.tika.OggDetector
　　org.apache.tika.mime.MimeTypes


现在我们该如何调用呢，


public static void main(String[] args) throws IOException {
// TODO Auto-generated method stub
ServiceLoader loader = new ServiceLoader();        
MimeTypes mimeTypes = MimeTypes.getDefaultMimeTypes();        
Detector detector=new DefaultDetector(mimeTypes, loader);      
File file=new File("[文件路径]");
InputStream stream = null;
try
{
stream=new BufferedInputStream(new FileInputStream(file));            
MediaType type =detector.detect(stream, new Metadata());
System.out.println("mime类型："+type.toString());
}
finally
{
if (stream != null)    stream.close();
}
}
　　现在还有tika怎样加载Parser实现类的，怎样根据文档的mime类型调用相应的Parser实现类的还没有进行分析，不过这些都相对容易分析了，下文再继续吧。