使用反射可以找到包中的所有类吗?

是否有可能找到给定包中的所有类或接口?(快速看一下例如Package,似乎不是。)

493032 次浏览

这是不可能的,因为包中的所有类都可能不会被加载,而你总是知道一个类的包。

由于类装入器的动态特性,这是不可能的。类装入器不需要告诉VM它可以提供哪些类,相反,它们只是提交类请求,并且必须返回类或抛出异常。

但是,如果您编写自己的类装入器,或者检查类路径和它的jar,就有可能找到这些信息。不过,这将通过文件系统操作,而不是反射。甚至可能有一些库可以帮助你做到这一点。

如果有远程生成或交付的类,您将无法发现这些类。

通常的方法是在某个文件中注册需要访问的类,或者在不同的类中引用它们。或者在命名时使用惯例。

反思图书馆将允许您在当前类路径中查找类。它可以用来获取包中的所有类:

 Reflections reflections = new Reflections("my.project.prefix");


Set<Class<? extends Object>> allClasses =
reflections.getSubTypesOf(Object.class);

您可以使用这个方法1,使用ClassLoader

/**
* Scans all classes accessible from the context class loader which belong to the given package and subpackages.
*
* @param packageName The base package
* @return The classes
* @throws ClassNotFoundException
* @throws IOException
*/
private static Class[] getClasses(String packageName)
throws ClassNotFoundException, IOException {
ClassLoader classLoader = Thread.currentThread().getContextClassLoader();
assert classLoader != null;
String path = packageName.replace('.', '/');
Enumeration<URL> resources = classLoader.getResources(path);
List<File> dirs = new ArrayList<File>();
while (resources.hasMoreElements()) {
URL resource = resources.nextElement();
dirs.add(new File(resource.getFile()));
}
ArrayList<Class> classes = new ArrayList<Class>();
for (File directory : dirs) {
classes.addAll(findClasses(directory, packageName));
}
return classes.toArray(new Class[classes.size()]);
}


/**
* Recursive method used to find all classes in a given directory and subdirs.
*
* @param directory   The base directory
* @param packageName The package name for classes found inside the base directory
* @return The classes
* @throws ClassNotFoundException
*/
private static List<Class> findClasses(File directory, String packageName) throws ClassNotFoundException {
List<Class> classes = new ArrayList<Class>();
if (!directory.exists()) {
return classes;
}
File[] files = directory.listFiles();
for (File file : files) {
if (file.isDirectory()) {
assert !file.getName().contains(".");
classes.addAll(findClasses(file, packageName + "." + file.getName()));
} else if (file.getName().endsWith(".class")) {
classes.add(Class.forName(packageName + '.' + file.getName().substring(0, file.getName().length() - 6)));
}
}
return classes;
}
< p >__________< br > 这个方法最初取自http://snippets.dzone.com/posts/show/4831,这是存档由互联网档案馆,现在链接。这个代码片段也可以在https://dzone.com/articles/get-all-classes-within-package.中找到

如果您不使用任何动态类加载器,您可以搜索类路径,并为每个条目搜索目录或JAR文件。

您可能应该看看开源的反映图书馆。有了它,你可以很容易地实现你想要的。

首先,设置反射索引(这有点乱,因为搜索所有类默认是禁用的):

List<ClassLoader> classLoadersList = new LinkedList<ClassLoader>();
classLoadersList.add(ClasspathHelper.contextClassLoader());
classLoadersList.add(ClasspathHelper.staticClassLoader());


Reflections reflections = new Reflections(new ConfigurationBuilder()
.setScanners(new SubTypesScanner(false /* don't exclude Object.class */), new ResourcesScanner())
.setUrls(ClasspathHelper.forClassLoader(classLoadersList.toArray(new ClassLoader[0])))
.filterInputsBy(new FilterBuilder().include(FilterBuilder.prefix("org.your.package"))));

然后你可以查询给定包中的所有对象:

Set<Class<?>> classes = reflections.getSubTypesOf(Object.class);

我整理了一个简单的github项目来解决这个问题:

https://github.com/ddopson/java-class-enumerator

它既适用于基于文件的类路径,也适用于jar文件。

如果你在签出项目后运行'make',它将打印出以下内容:

 Cleaning...
rm -rf build/
Building...
javac -d build/classes src/pro/ddopson/ClassEnumerator.java src/test/ClassIShouldFindOne.java src/test/ClassIShouldFindTwo.java src/test/subpkg/ClassIShouldFindThree.java src/test/TestClassEnumeration.java
Making JAR Files...
jar cf build/ClassEnumerator_test.jar -C build/classes/ .
jar cf build/ClassEnumerator.jar -C build/classes/ pro
Running Filesystem Classpath Test...
java -classpath build/classes test.TestClassEnumeration
ClassDiscovery: Package: 'test' becomes Resource: 'file:/Users/Dopson/work/other/java-class-enumeration/build/classes/test'
ClassDiscovery: Reading Directory '/Users/Dopson/work/other/java-class-enumeration/build/classes/test'
ClassDiscovery: FileName 'ClassIShouldFindOne.class'  =>  class 'test.ClassIShouldFindOne'
ClassDiscovery: FileName 'ClassIShouldFindTwo.class'  =>  class 'test.ClassIShouldFindTwo'
ClassDiscovery: FileName 'subpkg'  =>  class 'null'
ClassDiscovery: Reading Directory '/Users/Dopson/work/other/java-class-enumeration/build/classes/test/subpkg'
ClassDiscovery: FileName 'ClassIShouldFindThree.class'  =>  class 'test.subpkg.ClassIShouldFindThree'
ClassDiscovery: FileName 'TestClassEnumeration.class'  =>  class 'test.TestClassEnumeration'
Running JAR Classpath Test...
java -classpath build/ClassEnumerator_test.jar  test.TestClassEnumeration
ClassDiscovery: Package: 'test' becomes Resource: 'jar:file:/Users/Dopson/work/other/java-class-enumeration/build/ClassEnumerator_test.jar!/test'
ClassDiscovery: Reading JAR file: '/Users/Dopson/work/other/java-class-enumeration/build/ClassEnumerator_test.jar'
ClassDiscovery: JarEntry 'META-INF/'  =>  class 'null'
ClassDiscovery: JarEntry 'META-INF/MANIFEST.MF'  =>  class 'null'
ClassDiscovery: JarEntry 'pro/'  =>  class 'null'
ClassDiscovery: JarEntry 'pro/ddopson/'  =>  class 'null'
ClassDiscovery: JarEntry 'pro/ddopson/ClassEnumerator.class'  =>  class 'null'
ClassDiscovery: JarEntry 'test/'  =>  class 'null'
ClassDiscovery: JarEntry 'test/ClassIShouldFindOne.class'  =>  class 'test.ClassIShouldFindOne'
ClassDiscovery: JarEntry 'test/ClassIShouldFindTwo.class'  =>  class 'test.ClassIShouldFindTwo'
ClassDiscovery: JarEntry 'test/subpkg/'  =>  class 'null'
ClassDiscovery: JarEntry 'test/subpkg/ClassIShouldFindThree.class'  =>  class 'test.subpkg.ClassIShouldFindThree'
ClassDiscovery: JarEntry 'test/TestClassEnumeration.class'  =>  class 'test.TestClassEnumeration'
Tests Passed.

参见我的另一个答案

谷歌Guava 14包含了一个新类ClassPath,有三个方法来扫描顶级类:

  • # EYZ0
  • # EYZ0
  • # EYZ0

更多信息请参见# EYZ0 javadocs

一般来说,类装入器不允许扫描类路径上的所有类。但通常唯一使用的类加载器是UrlClassLoader,我们可以从中检索目录和jar文件的列表(参见geturl),并逐个打开它们以列出可用的类。这种方法称为类路径扫描,在Scannotation反射中实现。

Reflections reflections = new Reflections("my.package");
Set<Class<? extends Object>> classes = reflections.getSubTypesOf(Object.class);

另一种方法是使用Java可插入注释处理API编写注释处理器,它将在编译时收集所有注释类,并构建索引文件以供运行时使用。这个机制在ClassIndex库中实现:

// package-info.java
@IndexSubclasses
package my.package;


// your code
Iterable<Class> classes = ClassIndex.getPackageClasses("my.package");

注意,由于Java编译器自动发现类路径上的任何处理器,因此扫描是完全自动化的,因此不需要额外的设置。

你需要查找类路径中的每个类装入器条目:

    String pkg = "org/apache/commons/lang";
ClassLoader cl = ClassLoader.getSystemClassLoader();
URL[] urls = ((URLClassLoader) cl).getURLs();
for (URL url : urls) {
System.out.println(url.getFile());
File jar = new File(url.getFile());
// ....
}

如果条目是目录,只需在右边的子目录中查找:

if (jar.isDirectory()) {
File subdir = new File(jar, pkg);
if (!subdir.exists())
continue;
File[] files = subdir.listFiles();
for (File file : files) {
if (!file.isFile())
continue;
if (file.getName().endsWith(".class"))
System.out.println("Found class: "
+ file.getName().substring(0,
file.getName().length() - 6));
}
}

如果条目是文件,并且它是jar,检查它的ZIP条目:

else {
// try to open as ZIP
try {
ZipFile zip = new ZipFile(jar);
for (Enumeration<? extends ZipEntry> entries = zip
.entries(); entries.hasMoreElements();) {
ZipEntry entry = entries.nextElement();
String name = entry.getName();
if (!name.startsWith(pkg))
continue;
name = name.substring(pkg.length() + 1);
if (name.indexOf('/') < 0 && name.endsWith(".class"))
System.out.println("Found class: "
+ name.substring(0, name.length() - 6));
}
} catch (ZipException e) {
System.out.println("Not a ZIP: " + e.getMessage());
} catch (IOException e) {
System.err.println(e.getMessage());
}
}

现在,一旦包中有了所有的类名,就可以尝试用反射加载它们,并分析它们是类还是接口等等。

我一直在尝试使用Reflections库,但在使用它时遇到了一些问题,而且为了简单地获取包上的类,我应该包含太多的jar。

我将在这个重复的问题中找到一个解决方案:如何在包中获得所有类的名称?

# EYZ0;我添加了一些更正,使其工作:

import java.io.File;
import java.io.IOException;
import java.net.URL;
import java.util.Enumeration;
import java.util.LinkedList;
import java.util.List;


public final class ClassFinder {


private final static char DOT = '.';
private final static char SLASH = '/';
private final static String CLASS_SUFFIX = ".class";
private final static String BAD_PACKAGE_ERROR = "Unable to get resources from path '%s'. Are you sure the given '%s' package exists?";


public final static List<Class<?>> find(final String scannedPackage) {
final ClassLoader classLoader = Thread.currentThread().getContextClassLoader();
final String scannedPath = scannedPackage.replace(DOT, SLASH);
final Enumeration<URL> resources;
try {
resources = classLoader.getResources(scannedPath);
} catch (IOException e) {
throw new IllegalArgumentException(String.format(BAD_PACKAGE_ERROR, scannedPath, scannedPackage), e);
}
final List<Class<?>> classes = new LinkedList<Class<?>>();
while (resources.hasMoreElements()) {
final File file = new File(resources.nextElement().getFile());
classes.addAll(find(file, scannedPackage));
}
return classes;
}


private final static List<Class<?>> find(final File file, final String scannedPackage) {
final List<Class<?>> classes = new LinkedList<Class<?>>();
if (file.isDirectory()) {
for (File nestedFile : file.listFiles()) {
classes.addAll(find(nestedFile, scannedPackage));
}
//File names with the $1, $2 holds the anonymous inner classes, we are not interested on them.
} else if (file.getName().endsWith(CLASS_SUFFIX) && !file.getName().contains("$")) {


final int beginIndex = 0;
final int endIndex = file.getName().length() - CLASS_SUFFIX.length();
final String className = file.getName().substring(beginIndex, endIndex);
try {
final String resource = scannedPackage + DOT + className;
classes.add(Class.forName(resource));
} catch (ClassNotFoundException ignore) {
}
}
return classes;
}


}
要使用它,只需调用本例中提到的sp00n的find方法: 如果需要的话,我已经添加了类实例的创建
List<Class<?>> classes = ClassFinder.find("com.package");


ExcelReporting excelReporting;
for (Class<?> aClass : classes) {
Constructor constructor = aClass.getConstructor();
//Create an object of the class type
constructor.newInstance();
//...
}

我刚写了一个util类,里面包含了测试方法,大家可以检查一下~

IteratePackageUtil.java:

package eric.j2se.reflect;


import java.util.Set;


import org.reflections.Reflections;
import org.reflections.scanners.ResourcesScanner;
import org.reflections.scanners.SubTypesScanner;
import org.reflections.util.ClasspathHelper;
import org.reflections.util.ConfigurationBuilder;
import org.reflections.util.FilterBuilder;


/**
* an util to iterate class in a package,
*
* @author eric
* @date Dec 10, 2013 12:36:46 AM
*/
public class IteratePackageUtil {
/**
* <p>
* Get set of all class in a specified package recursively. this only support lib
* </p>
* <p>
* class of sub package will be included, inner class will be included,
* </p>
* <p>
* could load class that use the same classloader of current class, can't load system packages,
* </p>
*
* @param pkg
*            path of a package
* @return
*/
public static Set<Class<? extends Object>> getClazzSet(String pkg) {
// prepare reflection, include direct subclass of Object.class
Reflections reflections = new Reflections(new ConfigurationBuilder().setScanners(new SubTypesScanner(false), new ResourcesScanner())
.setUrls(ClasspathHelper.forClassLoader(ClasspathHelper.classLoaders(new ClassLoader[0])))
.filterInputsBy(new FilterBuilder().includePackage(pkg)));


return reflections.getSubTypesOf(Object.class);
}


public static void test() {
String pkg = "org.apache.tomcat.util";


Set<Class<? extends Object>> clazzSet = getClazzSet(pkg);
for (Class<? extends Object> clazz : clazzSet) {
System.out.println(clazz.getName());
}
}


public static void main(String[] args) {
test();
}
}

春天

这个例子是针对Spring 4的,但是您也可以在早期版本中找到类路径扫描器。

// create scanner and disable default filters (that is the 'false' argument)
final ClassPathScanningCandidateComponentProvider provider = new ClassPathScanningCandidateComponentProvider(false);
// add include filters which matches all the classes (or use your own)
provider.addIncludeFilter(new RegexPatternTypeFilter(Pattern.compile(".*")));


// get matching classes defined in the package
final Set<BeanDefinition> classes = provider.findCandidateComponents("my.package.name");


// this is how you can load the class type from BeanDefinition instance
for (BeanDefinition bean: classes) {
Class<?> clazz = Class.forName(bean.getBeanClassName());
// ... do your magic with the class ...
}

谷歌番石榴

在版本14中,API仍然被标记为@Beta,所以在生产代码中要小心。

final ClassLoader loader = Thread.currentThread().getContextClassLoader();


for (final ClassPath.ClassInfo info : ClassPath.from(loader).getTopLevelClasses()) {
if (info.getName().startsWith("my.package.")) {
final Class<?> clazz = info.load();
// do something with your clazz
}
}

不使用任何额外的库:

package test;


import java.io.DataInputStream;
import java.io.InputStream;
import java.net.URL;
import java.util.ArrayList;
import java.util.List;


public class Test {
public static void main(String[] args) throws Exception{
List<Class> classes = getClasses(Test.class.getClassLoader(),"test");
for(Class c:classes){
System.out.println("Class: "+c);
}
}


public static List<Class> getClasses(ClassLoader cl,String pack) throws Exception{


String dottedPackage = pack.replaceAll("[/]", ".");
List<Class> classes = new ArrayList<Class>();
URL upackage = cl.getResource(pack);


DataInputStream dis = new DataInputStream((InputStream) upackage.getContent());
String line = null;
while ((line = dis.readLine()) != null) {
if(line.endsWith(".class")) {
classes.add(Class.forName(dottedPackage+"."+line.substring(0,line.lastIndexOf('.'))));
}
}
return classes;
}
}

你好。我总是对上面的解决方案(和其他网站)有一些问题 我,作为一名开发人员,正在为API编程一个插件。该API防止使用任何外部库或第三方工具。设置还包括jar或zip文件中的代码和直接位于某些目录中的类文件。所以我的代码必须能够解决每一个设置。经过大量的研究,我想出了一个方法,可以在所有可能的设置中至少95%的工作

下面的代码基本上是总是有效的overkill方法。

代码:

这段代码扫描给定包中包含的所有类。它将只适用于当前ClassLoader中的所有类。

/**
* Private helper method
*
* @param directory
*            The directory to start with
* @param pckgname
*            The package name to search for. Will be needed for getting the
*            Class object.
* @param classes
*            if a file isn't loaded but still is in the directory
* @throws ClassNotFoundException
*/
private static void checkDirectory(File directory, String pckgname,
ArrayList<Class<?>> classes) throws ClassNotFoundException {
File tmpDirectory;


if (directory.exists() && directory.isDirectory()) {
final String[] files = directory.list();


for (final String file : files) {
if (file.endsWith(".class")) {
try {
classes.add(Class.forName(pckgname + '.'
+ file.substring(0, file.length() - 6)));
} catch (final NoClassDefFoundError e) {
// do nothing. this class hasn't been found by the
// loader, and we don't care.
}
} else if ((tmpDirectory = new File(directory, file))
.isDirectory()) {
checkDirectory(tmpDirectory, pckgname + "." + file, classes);
}
}
}
}


/**
* Private helper method.
*
* @param connection
*            the connection to the jar
* @param pckgname
*            the package name to search for
* @param classes
*            the current ArrayList of all classes. This method will simply
*            add new classes.
* @throws ClassNotFoundException
*             if a file isn't loaded but still is in the jar file
* @throws IOException
*             if it can't correctly read from the jar file.
*/
private static void checkJarFile(JarURLConnection connection,
String pckgname, ArrayList<Class<?>> classes)
throws ClassNotFoundException, IOException {
final JarFile jarFile = connection.getJarFile();
final Enumeration<JarEntry> entries = jarFile.entries();
String name;


for (JarEntry jarEntry = null; entries.hasMoreElements()
&& ((jarEntry = entries.nextElement()) != null);) {
name = jarEntry.getName();


if (name.contains(".class")) {
name = name.substring(0, name.length() - 6).replace('/', '.');


if (name.contains(pckgname)) {
classes.add(Class.forName(name));
}
}
}
}


/**
* Attempts to list all the classes in the specified package as determined
* by the context class loader
*
* @param pckgname
*            the package name to search
* @return a list of classes that exist within that package
* @throws ClassNotFoundException
*             if something went wrong
*/
public static ArrayList<Class<?>> getClassesForPackage(String pckgname)
throws ClassNotFoundException {
final ArrayList<Class<?>> classes = new ArrayList<Class<?>>();


try {
final ClassLoader cld = Thread.currentThread()
.getContextClassLoader();


if (cld == null)
throw new ClassNotFoundException("Can't get class loader.");


final Enumeration<URL> resources = cld.getResources(pckgname
.replace('.', '/'));
URLConnection connection;


for (URL url = null; resources.hasMoreElements()
&& ((url = resources.nextElement()) != null);) {
try {
connection = url.openConnection();


if (connection instanceof JarURLConnection) {
checkJarFile((JarURLConnection) connection, pckgname,
classes);
} else if (connection instanceof FileURLConnection) {
try {
checkDirectory(
new File(URLDecoder.decode(url.getPath(),
"UTF-8")), pckgname, classes);
} catch (final UnsupportedEncodingException ex) {
throw new ClassNotFoundException(
pckgname
+ " does not appear to be a valid package (Unsupported encoding)",
ex);
}
} else
throw new ClassNotFoundException(pckgname + " ("
+ url.getPath()
+ ") does not appear to be a valid package");
} catch (final IOException ioex) {
throw new ClassNotFoundException(
"IOException was thrown when trying to get all resources for "
+ pckgname, ioex);
}
}
} catch (final NullPointerException ex) {
throw new ClassNotFoundException(
pckgname
+ " does not appear to be a valid package (Null pointer exception)",
ex);
} catch (final IOException ioex) {
throw new ClassNotFoundException(
"IOException was thrown when trying to get all resources for "
+ pckgname, ioex);
}


return classes;
}

这三个方法为你提供了在给定的包中找到所有类的能力 你可以这样使用它:

getClassesForPackage("package.your.classes.are.in");

解释:

该方法首先获取当前的ClassLoader。然后它获取包含该包的所有资源,并迭代这些URLs。然后它创建一个URLConnection并确定我们拥有的URl类型。它可以是一个目录(FileURLConnection),也可以是一个jar或zip文件中的目录(JarURLConnection)。根据我们所拥有的连接类型,将调用两个不同的方法。

首先让我们看看如果它是FileURLConnection会发生什么 它首先检查传递的File是否存在,是否为目录。如果是这样,它会检查它是否是类文件。如果是这样,一个Class对象将被创建并放在ArrayList中。如果它不是一个类文件,而是一个目录,我们只需迭代它,并做同样的事情。所有其他情况/文件将被忽略

如果URLConnectionJarURLConnection,则另一个私有helper方法将被调用。此方法迭代zip/jar存档中的所有条目。如果一个条目是类文件,并且位于包内部,则将创建一个Class对象并将其存储在ArrayList中。

在解析完所有资源后,它(主方法)返回包含给定包中当前ClassLoader所知道的所有类的ArrayList

如果进程在任何地方失败,将抛出一个ClassNotFoundException,其中包含有关确切原因的详细信息。

基于@Staale的回答,并试图不依赖于第三方库,我将通过检查第一个包的物理位置来实现文件系统方法:

import java.io.File;
import java.io.FileFilter;
import java.util.ArrayList;
...
Class<?>[] foundClasses = new Class<?>[0];
final ArrayList<Class<?>> foundClassesDyn = new ArrayList<Class<?>>();


new java.io.File(
klass.getResource(
"/" + curPackage.replace( "." , "/")
).getFile()
).listFiles(
new java.io.FileFilter() {
public boolean accept(java.io.File file) {
final String classExtension = ".class";


if ( file.isFile()
&& file.getName().endsWith(classExtension)
// avoid inner classes
&& ! file.getName().contains("$") )
{
try {
String className = file.getName();
className = className.substring(0, className.length() - classExtension.length());
foundClassesDyn.add( Class.forName( curPackage + "." + className ) );
} catch (ClassNotFoundException e) {
e.printStackTrace(System.out);
}
}


return false;
}
}
);


foundClasses = foundClassesDyn.toArray(foundClasses);

几乎所有的答案都使用Reflections或从文件系统读取类文件。如果尝试从文件系统读取类,则在将应用程序打包为JAR或其他格式时可能会出错。此外,您可能不想为此目的使用单独的库。

这里有另一种方法,它是纯java,不依赖于文件系统。

import javax.tools.JavaFileObject;
import javax.tools.StandardJavaFileManager;
import javax.tools.StandardLocation;
import javax.tools.ToolProvider;
import java.io.File;
import java.util.ArrayList;
import java.util.Collection;
import java.util.Collections;
import java.util.regex.Pattern;
import java.util.stream.Collectors;
import java.util.stream.StreamSupport;


public class PackageUtil {


public static Collection<Class> getClasses(final String pack) throws Exception {
final StandardJavaFileManager fileManager = ToolProvider.getSystemJavaCompiler().getStandardFileManager(null, null, null);
return StreamSupport.stream(fileManager.list(StandardLocation.CLASS_PATH, pack, Collections.singleton(JavaFileObject.Kind.CLASS), false).spliterator(), false)
.map(javaFileObject -> {
try {
final String[] split = javaFileObject.getName()
.replace(".class", "")
.replace(")", "")
.split(Pattern.quote(File.separator));


final String fullClassName = pack + "." + split[split.length - 1];
return Class.forName(fullClassName);
} catch (ClassNotFoundException e) {
throw new RuntimeException(e);
}


})
.collect(Collectors.toCollection(ArrayList::new));
}
}
< p > # EYZ0。你可以使用for循环代替流。 你可以这样测试

public static void main(String[] args) throws Exception {
final String pack = "java.nio.file"; // Or any other package
PackageUtil.getClasses(pack).stream().forEach(System.out::println);
}

值得一提的

如果你想要一个包下所有类的列表,你可以用下面的方式使用Reflection:

List<Class> myTypes = new ArrayList<>();


Reflections reflections = new Reflections("com.package");
for (String s : reflections.getStore().get(SubTypesScanner.class).values()) {
myTypes.add(Class.forName(s));
}

这将创建一个类列表,稍后您可以随心所欲地使用它们。

目前列出给定包中所有类的最健壮的机制是ClassGraph,因为它处理尽可能广泛的类路径规范机制,包括新的JPMS模块系统。(我是作者。)

List<String> classNames = new ArrayList<>();
try (ScanResult scanResult = new ClassGraph().acceptPackages("my.package")
.enableClassInfo().scan()) {
classNames.addAll(scanResult.getAllClasses().getNames());
}

我是这样做的。我扫描所有的子文件夹(子包),我不尝试加载匿名类:

   /**
* Attempts to list all the classes in the specified package as determined
* by the context class loader, recursively, avoiding anonymous classes
*
* @param pckgname
*            the package name to search
* @return a list of classes that exist within that package
* @throws ClassNotFoundException
*             if something went wrong
*/
private static List<Class> getClassesForPackage(String pckgname) throws ClassNotFoundException {
// This will hold a list of directories matching the pckgname. There may be more than one if a package is split over multiple jars/paths
ArrayList<File> directories = new ArrayList<File>();
String packageToPath = pckgname.replace('.', '/');
try {
ClassLoader cld = Thread.currentThread().getContextClassLoader();
if (cld == null) {
throw new ClassNotFoundException("Can't get class loader.");
}


// Ask for all resources for the packageToPath
Enumeration<URL> resources = cld.getResources(packageToPath);
while (resources.hasMoreElements()) {
directories.add(new File(URLDecoder.decode(resources.nextElement().getPath(), "UTF-8")));
}
} catch (NullPointerException x) {
throw new ClassNotFoundException(pckgname + " does not appear to be a valid package (Null pointer exception)");
} catch (UnsupportedEncodingException encex) {
throw new ClassNotFoundException(pckgname + " does not appear to be a valid package (Unsupported encoding)");
} catch (IOException ioex) {
throw new ClassNotFoundException("IOException was thrown when trying to get all resources for " + pckgname);
}


ArrayList<Class> classes = new ArrayList<Class>();
// For every directoryFile identified capture all the .class files
while (!directories.isEmpty()){
File directoryFile  = directories.remove(0);
if (directoryFile.exists()) {
// Get the list of the files contained in the package
File[] files = directoryFile.listFiles();


for (File file : files) {
// we are only interested in .class files
if ((file.getName().endsWith(".class")) && (!file.getName().contains("$"))) {
// removes the .class extension
int index = directoryFile.getPath().indexOf(packageToPath);
String packagePrefix = directoryFile.getPath().substring(index).replace('/', '.');;
try {
String className = packagePrefix + '.' + file.getName().substring(0, file.getName().length() - 6);
classes.add(Class.forName(className));
} catch (NoClassDefFoundError e)
{
// do nothing. this class hasn't been found by the loader, and we don't care.
}
} else if (file.isDirectory()){ // If we got to a subdirectory
directories.add(new File(file.getPath()));
}
}
} else {
throw new ClassNotFoundException(pckgname + " (" + directoryFile.getPath() + ") does not appear to be a valid package");
}
}
return classes;
}

是的,你可以使用很少的API,这是我喜欢做的,面对这个问题,我使用hibernate核心&必须找到带有特定注释的类。

使这些自定义注释使用,您将标记哪些类您想要获得。

import java.lang.annotation.ElementType;
import java.lang.annotation.Retention;
import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;


@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.TYPE)
public @interface EntityToBeScanned {


}

然后用它来标记你的课

@EntityToBeScanned
public MyClass{


}

创建具有以下方法的实用程序类

public class ClassScanner {


public static Set<Class<?>> allFoundClassesAnnotatedWithEntityToBeScanned(){
Reflections reflections = new Reflections(".*");
Set<Class<?>> annotated = reflections.getTypesAnnotatedWith(EntityToBeScanned.class);
return annotated;
}


}

调用allFoundClassesAnnotatedWithEntityToBeScanned ()方法获取已找到的类的

你将需要下列的参考书目

<!-- https://mvnrepository.com/artifact/com.google.guava/guava -->
<dependency>
<groupId>com.google.guava</groupId>
<artifactId>guava</artifactId>
<version>21.0</version>
</dependency>
<!-- https://mvnrepository.com/artifact/org.javassist/javassist -->
<dependency>
<groupId>org.javassist</groupId>
<artifactId>javassist</artifactId>
<version>3.22.0-CR1</version>
</dependency>
<!-- https://mvnrepository.com/artifact/org.reflections/reflections -->
<dependency>
<groupId>org.reflections</groupId>
<artifactId>reflections</artifactId>
<version>0.9.10</version>
</dependency>
这是非常可能的,但是如果没有像Reflections这样的附加库,这是很难的 这是困难的,因为你没有完整的工具获取类名 并且,我取我的ClassFinder类的代码:

package play.util;


import java.io.File;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Enumeration;
import java.util.List;
import java.util.jar.JarEntry;
import java.util.jar.JarFile;


/**
* Created by LINKOR on 26.05.2017 in 15:12.
* Date: 2017.05.26
*/
public class FileClassFinder {
private JarFile file;
private boolean trouble;
public FileClassFinder(String filePath) {
try {
file = new JarFile(filePath);
} catch (IOException e) {
trouble = true;
}
}


public List<String> findClasses(String pkg) {
ArrayList<String> classes = new ArrayList<>();
Enumeration<JarEntry> entries = file.entries();
while (entries.hasMoreElements()) {
JarEntry cls = entries.nextElement();
if (!cls.isDirectory()) {
String fileName = cls.getName();
String className = fileName.replaceAll("/",         ".").replaceAll(File.pathSeparator, ".").substring(0, fileName.lastIndexOf('.'));
if (className.startsWith(pkg)) classes.add(className.substring(pkg.length() + 1));
}
}
return classes;
}
}

在使用Maven时,Aleksander Blomskøld的解决方案对于参数化测试@RunWith(Parameterized.class)不适用。正确命名了测试,并显示了找到但未执行的位置:

-------------------------------------------------------
T E S T S
-------------------------------------------------------
Running some.properly.named.test.run.with.maven.SomeTest
Tests run: 0, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.123 sec

在这里报告了一个类似的问题。

在我的例子中,@Parameters是在包中创建每个类的实例。这些测试在IDE中本地运行时运行良好。然而,当运行Maven时,在Aleksander Blomskøld的解决方案中没有找到类。

我确实用以下剪辑完成了它,灵感来自David Pärsson对Aleksander Blomskøld回答的评论:

Reflections reflections = new Reflections(new ConfigurationBuilder()
.setScanners(new SubTypesScanner(false /* don't exclude Object.class */), new ResourcesScanner())
.addUrls(ClasspathHelper.forJavaClassPath())
.filterInputsBy(new FilterBuilder()
.include(FilterBuilder.prefix(basePackage))));


Set<Class<?>> subTypesOf = reflections.getSubTypesOf(Object.class);

如果您只是想加载一组相关的类,那么Spring可以帮助您。

Spring可以在一行代码中实例化实现给定接口的所有类的列表或映射。列表或映射将包含实现该接口的所有类的实例。

也就是说,作为从文件系统中加载类列表的替代方法,只需在您想要加载的所有类中实现相同的接口,而不管包是什么,并使用Spring为您提供所有这些类的实例。这样,您就可以加载(并实例化)您想要的所有类,而不管它们在哪个包中。

另一方面,如果您希望将它们都放在一个包中,那么只需让包中的所有类实现给定的接口。

注意,接口本身不需要声明任何方法——它可以完全为空。

要注入实现给定接口的类列表,请使用以下代码行…

  @Autowired
private List<ISomeInterface> implementationList;

也可以使用Spring注入类的Map。如果有兴趣,请阅读文档。

最后,我将提供一个比搜索整个文件系统树更优雅的解决方案。

创建一个自定义注释,用于构建应用于它的类的目录——类似于@ClassCatalog。

plain java: FindAllClassesUsingPlainJavaReflectionTest.java

@Slf4j
class FindAllClassesUsingPlainJavaReflectionTest {


private static final Function<Throwable, RuntimeException> asRuntimeException = throwable -> {
log.error(throwable.getLocalizedMessage());
return new RuntimeException(throwable);
};


private static final Function<String, Collection<Class<?>>> findAllPackageClasses = basePackageName -> {


Locale locale = Locale.getDefault();
Charset charset = StandardCharsets.UTF_8;
val fileManager = ToolProvider.getSystemJavaCompiler()
.getStandardFileManager(/* diagnosticListener */ null, locale, charset);


StandardLocation location = StandardLocation.CLASS_PATH;
JavaFileObject.Kind kind = JavaFileObject.Kind.CLASS;
Set<JavaFileObject.Kind> kinds = Collections.singleton(kind);
val javaFileObjects = Try.of(() -> fileManager.list(location, basePackageName, kinds, /* recurse */ true))
.getOrElseThrow(asRuntimeException);


String pathToPackageAndClass = basePackageName.replace(".", File.separator);
Function<String, String> mapToClassName = s -> {
String prefix = Arrays.stream(s.split(pathToPackageAndClass))
.findFirst()
.orElse("");
return s.replaceFirst(prefix, "")
.replaceAll(File.separator, ".");
};


return StreamSupport.stream(javaFileObjects.spliterator(), /* parallel */ true)
.filter(javaFileObject -> javaFileObject.getKind().equals(kind))
.map(FileObject::getName)
.map(fileObjectName -> fileObjectName.replace(".class", ""))
.map(mapToClassName)
.map(className -> Try.of(() -> Class.forName(className))
.getOrElseThrow(asRuntimeException))
.collect(Collectors.toList());
};


@Test
@DisplayName("should get classes recursively in given package")
void test() {
Collection<Class<?>> classes = findAllPackageClasses.apply(getClass().getPackage().getName());
assertThat(classes).hasSizeGreaterThan(4);
classes.stream().map(String::valueOf).forEach(log::info);
}
}

PS:为了简化处理错误的样板,我在这里使用vavrlombok

其他实现可以在我的GitHub daggerok/java-reflection-find- annotation -classes-or-methods repo中找到

那么这个呢:

public static List<Class<?>> getClassesForPackage(final String pkgName) throws IOException, URISyntaxException {
final String pkgPath = pkgName.replace('.', '/');
final URI pkg = Objects.requireNonNull(ClassLoader.getSystemClassLoader().getResource(pkgPath)).toURI();
final ArrayList<Class<?>> allClasses = new ArrayList<Class<?>>();


Path root;
if (pkg.toString().startsWith("jar:")) {
try {
root = FileSystems.getFileSystem(pkg).getPath(pkgPath);
} catch (final FileSystemNotFoundException e) {
root = FileSystems.newFileSystem(pkg, Collections.emptyMap()).getPath(pkgPath);
}
} else {
root = Paths.get(pkg);
}


final String extension = ".class";
try (final Stream<Path> allPaths = Files.walk(root)) {
allPaths.filter(Files::isRegularFile).forEach(file -> {
try {
final String path = file.toString().replace('/', '.');
final String name = path.substring(path.indexOf(pkgName), path.length() - extension.length());
allClasses.add(Class.forName(name));
} catch (final ClassNotFoundException | StringIndexOutOfBoundsException ignored) {
}
});
}
return allClasses;
}

然后你可以重载这个函数:

public static List<Class<?>> getClassesForPackage(final Package pkg) throws IOException, URISyntaxException {
return getClassesForPackage(pkg.getName());
}

如果你需要测试:

public static void main(final String[] argv) throws IOException, URISyntaxException {
for (final Class<?> cls : getClassesForPackage("my.package")) {
System.out.println(cls);
}
for (final Class<?> cls : getClassesForPackage(MyClass.class.getPackage())) {
System.out.println(cls);
}
}

如果你的IDE没有导入helper:

import java.io.IOException;
import java.net.URI;
import java.net.URISyntaxException;
import java.nio.file.FileSystemNotFoundException;
import java.nio.file.FileSystems;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
import java.util.Objects;
import java.util.stream.Stream;

工作原理:

  • 从您的IDE

  • JAR文件

  • 没有外部依赖

我无法为如此简单的东西找到一个简短的工作片段。所以在这里,我自己做了一段时间后:

Reflections reflections =
new Reflections(new ConfigurationBuilder()
.filterInputsBy(new FilterBuilder().includePackage(packagePath))
.setUrls(ClasspathHelper.forPackage(packagePath))
.setScanners(new SubTypesScanner(false)));


Set<String> typeList = reflections.getAllTypes();

它使用org.reflections

如果你在Spring-land,你可以使用PathMatchingResourcePatternResolver;

  PathMatchingResourcePatternResolver resolver = new PathMatchingResourcePatternResolver();
Resource[] resources = resolver.getResources("classpath*:some/package/name/*.class");


Arrays.asList(resources).forEach(r->{
...
});

org.reflections版本0.10开始:

org.reflections.scanners.SubTypesScanner

而且

org.reflections.Reflections.getAllTypes()

弃用。我userd:

public Set<String> getEntityNamesInPackage(String packagePath) {
Reflections reflections = new Reflections(new ConfigurationBuilder()
.filterInputsBy(new FilterBuilder().includePackage(packagePath))
.setUrls(ClasspathHelper.forPackage(packagePath))
.setScanners(SubTypes.filterResultsBy(s -> true)));


return reflections.getAll(SubTypes).stream()
.filter(s -> s.startsWith(packagePath))
.collect(Collectors.toSet());
}

扫描类加载器和所有父加载器的jar文件和目录。 jar文件和由jar的Class-Path引用的目录也会被加载

这段代码是用Java 8、11、18测试的。 在8上,使用URLClassLoader和getURLs()方法,一切都完美地工作。 在11上,它使用反射工作得很好,但是JVM在stderr流上打印一个警告(不能用System.setErr()与我的JVM重定向) 在18反射是无用的(抛出NoSuchMethod/Field),唯一的事情(我知道它工作的地方)是使用getResource()方法。当类装入器从文件系统装入给定包的资源时,返回一个简单的路径url。当类加载器从jar中加载资源时,url像'jar:file:[jar-path]!

. /[in-jar-path]

我使用了答案https://stackoverflow.com/a/1157352/18252455(来自一个重复的问题),并添加了读取类路径和搜索目录url的功能。

/**
* orig description:<br>
* Scans all classloaders for the current thread for loaded jars, and then scans
* each jar for the package name in question, listing all classes directly under
* the package name in question. Assumes directory structure in jar file and class
* package naming follow java conventions (i.e. com.example.test.MyTest would be in
* /com/example/test/MyTest.class)
* <p>
* in addition this method also scans for directories, where also is assumed, that the classes are
* placed followed by the java conventions. (i.e. <code>com.example.test.MyTest</code> would be in
* <code>directory/com/example/test/MyTest.class</code>)
* <p>
* this method also reads the jars Class-Path for other jars and directories. for the jars and
* directories referred in the jars are scanned with the same rules as defined here.<br>
* it is ensured that no jar/directory is scanned exactly one time.
* <p>
* if {@code bailError} is <code>true</code> all errors will be wrapped in a
* {@link RuntimeException}
* and then thrown.<br>
* a {@link RuntimeException} will also be thrown if something unexpected happens.<br>
*
* @param packageName
*            the name of the package for which the classes should be searched
* @param allowSubPackages
*            <code>true</code> is also classes in sub packages should be found
* @param loader
*            the {@link ClassLoader} which should be used to find the URLs and to load classes
* @param bailError
*            if all {@link Exception} should be re-thrown wrapped in {@link RuntimeException} and
*            if a {@link RuntimeException} should be thrown, when something is not as expected.
* @see https://stackoverflow.com/questions/1156552/java-package-introspection
* @see https://stackoverflow.com/a/1157352/18252455
* @see https://creativecommons.org/licenses/by-sa/2.5/
* @see https://creativecommons.org/licenses/by-sa/2.5/legalcode
*/
public static Set <Class <?>> tryGetClassesForPackage(String packageName, boolean allowSubPackages, ClassLoader loader, boolean bailError) {
Set <URL> jarUrls = new HashSet <URL>();
Set <Path> directorys = new HashSet <Path>();
findClassPools(loader, jarUrls, directorys, bailError, packageName);
Set <Class <?>> jarClasses = findJarClasses(allowSubPackages, packageName, jarUrls, directorys, loader, bailError);
Set <Class <?>> dirClasses = findDirClasses(allowSubPackages, packageName, directorys, loader, bailError);
jarClasses.addAll(dirClasses);
return jarClasses;
}


private static Set <Class <?>> findDirClasses(boolean subPackages, String packageName, Set <Path> directorys, ClassLoader loader, boolean bailError) {
Filter <Path> filter;
Set <Class <?>> result = new HashSet <>();
for (Path startPath : directorys) {
String packagePath = packageName.replace(".", startPath.getFileSystem().getSeparator());
final Path searchPath = startPath.resolve(packagePath).toAbsolutePath();
if (subPackages) {
filter = p -> {
p = p.toAbsolutePath();
Path other;
if (p.getNameCount() >= searchPath.getNameCount()) {
other = searchPath;
} else {
other = searchPath.subpath(0, p.getNameCount());
}
if (p.startsWith(other)) {
return true;
} else {
return false;
}
};
} else {
filter = p -> {
p = p.toAbsolutePath();
if (p.getNameCount() > searchPath.getNameCount() + 1) {
return false;
} else if (p.toAbsolutePath().startsWith(searchPath)) {
return true;
} else {
return false;
}
};
}
if (Files.exists(searchPath)) {
findDirClassFilesRecursive(filter, searchPath, startPath, result, loader, bailError);
} // the package does not have to exist in every directory
}
return result;
}


private static void findDirClassFilesRecursive(Filter <Path> filter, Path path, Path start, Set <Class <?>> classes, ClassLoader loader, boolean bailError) {
try (DirectoryStream <Path> dirStream = Files.newDirectoryStream(path, filter)) {
for (Path p : dirStream) {
if (Files.isDirectory(p)) {
findDirClassFilesRecursive(filter, p, start, classes, loader, bailError);
} else {
Path subp = p.subpath(start.getNameCount(), p.getNameCount());
String str = subp.toString();
if (str.endsWith(".class")) {
str = str.substring(0, str.length() - 6);
String sep = p.getFileSystem().getSeparator();
if (str.startsWith(sep)) {
str = str.substring(sep.length());
}
if (str.endsWith(sep)) {
str = str.substring(0, str.length() - sep.length());
}
String fullClassName = str.replace(sep, ".");
try {
Class <?> cls = Class.forName(fullClassName, false, loader);
classes.add(cls);
} catch (ClassNotFoundException e) {
if (bailError) {
throw new RuntimeException(e);
}
}
}
}
}
} catch (IOException e) {
if (bailError) {
throw new RuntimeException(e);
}
}
}


private static Set <Class <?>> findJarClasses(boolean subPackages, String packageName, Set <URL> nextJarUrls, Set <Path> directories, ClassLoader loader, boolean bailError) {
String packagePath = packageName.replace('.', '/');
Set <Class <?>> result = new HashSet <>();
Set <URL> allJarUrls = new HashSet <>();
while (true) {
Set <URL> thisJarUrls = new HashSet <>(nextJarUrls);
thisJarUrls.removeAll(allJarUrls);
if (thisJarUrls.isEmpty()) {
break;
}
allJarUrls.addAll(thisJarUrls);
for (URL url : thisJarUrls) {
try (JarInputStream stream = new JarInputStream(url.openStream())) {
// may want better way to open url connections
readJarClassPath(stream, nextJarUrls, directories, bailError);
JarEntry entry = stream.getNextJarEntry();
while (entry != null) {
String name = entry.getName();
int i = name.lastIndexOf("/");
                    

if (i > 0 && name.endsWith(".class")) {
try {
if (subPackages) {
if (name.substring(0, i).startsWith(packagePath)) {
Class <?> cls = Class.forName(name.substring(0, name.length() - 6).replace("/", "."), false, loader);
result.add(cls);
}
} else {
if (name.substring(0, i).equals(packagePath)) {
Class <?> cls = Class.forName(name.substring(0, name.length() - 6).replace("/", "."), false, loader);
result.add(cls);
}
}
} catch (ClassNotFoundException e) {
e.printStackTrace();
}
}
entry = stream.getNextJarEntry();
}
stream.close();
} catch (IOException e) {
e.printStackTrace();
}
}
}
return result;
}


private static void readJarClassPath(JarInputStream stream, Set <URL> jarUrls, Set <Path> directories, boolean bailError) {
Object classPathObj = stream.getManifest().getMainAttributes().get(new Name("Class-Path"));
if (classPathObj == null) {
return;
}
if (classPathObj instanceof String) {
String[] entries = ((String) classPathObj).split("\\s+");// should also work with a single space (" ")
for (String entry : entries) {
try {
URL url = new URL(entry);
addFromUrl(jarUrls, directories, url, bailError);
} catch (MalformedURLException e) {
e.printStackTrace();
}
}
} else if (bailError) {
throw new RuntimeException("the Class-Path attribute is no String: " + classPathObj.getClass().getName() + " tos='" + classPathObj + "'");
}
}


private static void findClassPools(ClassLoader classLoader, Set <URL> jarUrls, Set <Path> directoryPaths, boolean bailError, String packageName) {
packageName = packageName.replace('.', '/');
while (classLoader != null) {
if (classLoader instanceof URLClassLoader) {
for (URL url : ((URLClassLoader) classLoader).getURLs()) {
addFromUrl(jarUrls, directoryPaths, url, bailError);
System.out.println("rurl-class-loade.url[n]r->'" + url + "'");
}
} else {
URL res = classLoader.getResource("");
if (res != null) {
addFromUrl(jarUrls, directoryPaths, res, bailError);
}
res = classLoader.getResource("/");
if (res != null) {
addFromUrl(jarUrls, directoryPaths, res, bailError);
}
res = classLoader.getResource("/" + packageName);
if (res != null) {
res = removePackageFromUrl(res, packageName, bailError);
if (res != null) {
addFromUrl(jarUrls, directoryPaths, res, bailError);
}
}
res = classLoader.getResource(packageName);
if (res != null) {
res = removePackageFromUrl(res, packageName, bailError);
if (res != null) {
addFromUrl(jarUrls, directoryPaths, res, bailError);
}
}
addFromUnknownClass(classLoader, jarUrls, directoryPaths, bailError, 8);
}
classLoader = classLoader.getParent();
}
}


private static URL removePackageFromUrl(URL res, String packagePath, boolean bailError) {
packagePath = "/" + packagePath;
String urlStr = res.toString();
if ( !urlStr.endsWith(packagePath)) {
if (bailError) {
throw new RuntimeException("the url string does not end with the packagepath! packagePath='" + packagePath + "' urlStr='" + urlStr + "'");
} else {
return null;
}
}
urlStr = urlStr.substring(0, urlStr.length() - packagePath.length());
if (urlStr.endsWith("!")) {
urlStr = urlStr.substring(0, urlStr.length() - 1);
}
if (urlStr.startsWith("jar:")) {
urlStr = urlStr.substring(4);
}
try {
return new URL(urlStr);
} catch (MalformedURLException e) {
if (bailError) {
throw new RuntimeException(e);
} else {
return null;
}
}
}


private static void addFromUnknownClass(Object instance, Set <URL> jarUrls, Set <Path> directoryPaths, boolean bailError, int maxDeep) {
Class <?> cls = instance.getClass();
while (cls != null) {
Field[] fields = cls.getDeclaredFields();
for (Field field : fields) {
Class <?> type = field.getType();
Object value;
try {
value = getValue(instance, field);
if (value != null) {
addFromUnknownValue(value, jarUrls, directoryPaths, bailError, type, field.getName(), maxDeep - 1);
}
} catch (IllegalArgumentException | IllegalAccessException | SecurityException e) {
if (bailError) {
final String version = System.getProperty("java.version");
String vers = version;
if (vers.startsWith("1.")) {
vers = vers.substring(2);
}
int dotindex = vers.indexOf('.');
if (dotindex != -1) {
vers = vers.substring(0, dotindex);
}
int versNum;
try {
versNum = Integer.parseInt(vers);
} catch (NumberFormatException e1) {
throw new RuntimeException("illegal version: '" + version + "' lastError: " + e.getMessage(), e);
}
if (versNum <= 11) {
throw new RuntimeException(e);
}
}
}
}
cls = cls.getSuperclass();
}
    

}


private static Object getValue(Object instance, Field field) throws IllegalArgumentException, IllegalAccessException, SecurityException {
try {
boolean flag = field.isAccessible();
boolean newflag = flag;
try {
field.setAccessible(true);
newflag = true;
} catch (Exception e) {}
try {
return field.get(instance);
} finally {
if (flag != newflag) {
field.setAccessible(flag);
}
}
} catch (IllegalArgumentException | IllegalAccessException | SecurityException e) {
try {
Field override = AccessibleObject.class.getDeclaredField("override");
boolean flag = override.isAccessible();
boolean newFlag = flag;
try {
override.setAccessible(true);
flag = true;
} catch (Exception s) {}
override.setBoolean(field, true);
if (flag != newFlag) {
override.setAccessible(flag);
}
return field.get(instance);
} catch (IllegalArgumentException | IllegalAccessException | NoSuchFieldException | SecurityException e1) {
e.addSuppressed(e1);
throw e;
}
}
}


private static void addFromUnknownValue(Object value, Set <URL> jarUrls, Set <Path> directoryPaths, boolean bailError, Class <?> type, String fieldName, int maxDeep) {
if (Collection.class.isAssignableFrom(type)) {
for (Object obj : (Collection <?>) value) {
URL url = null;
try {
if (obj instanceof URL) {
url = (URL) obj;
} else if (obj instanceof Path) {
url = ((Path) obj).toUri().toURL();
} else if (obj instanceof File) {
url = ((File) obj).toURI().toURL();
}
} catch (MalformedURLException e) {
if (bailError) {
throw new RuntimeException(e);
}
}
if (url != null) {
addFromUrl(jarUrls, directoryPaths, url, bailError);
}
}
} else if (URL[].class.isAssignableFrom(type)) {
for (URL url : (URL[]) value) {
addFromUrl(jarUrls, directoryPaths, url, bailError);
}
} else if (Path[].class.isAssignableFrom(type)) {
for (Path path : (Path[]) value) {
try {
addFromUrl(jarUrls, directoryPaths, path.toUri().toURL(), bailError);
} catch (MalformedURLException e) {
if (bailError) {
throw new RuntimeException(e);
}
}
}
} else if (File[].class.isAssignableFrom(type)) {
for (File file : (File[]) value) {
try {
addFromUrl(jarUrls, directoryPaths, file.toURI().toURL(), bailError);
} catch (MalformedURLException e) {
if (bailError) {
throw new RuntimeException(e);
}
}
}
} else if (maxDeep > 0) {
addFromUnknownClass(value, jarUrls, directoryPaths, bailError, maxDeep - 1);
}
}


private static void addFromUrl(Set <URL> jarUrls, Set <Path> directoryPaths, URL url, boolean bailError) {
if (url.getFile().endsWith(".jar") || url.getFile().endsWith(".zip")) {
// may want better way to detect jar files
jarUrls.add(url);
} else {
try {
Path path = Paths.get(url.toURI());
if (Files.isDirectory(path)) {
directoryPaths.add(path);
} else if (bailError) {
throw new RuntimeException("unknown url for class loading: " + url);
}
} catch (URISyntaxException e) {
if (bailError) {
throw new RuntimeException(e);
}
}
}
}

进口:

import java.io.File;
import java.io.IOException;
import java.lang.reflect.AccessibleObject;
import java.lang.reflect.Field;
import java.net.MalformedURLException;
import java.net.URISyntaxException;
import java.net.URL;
import java.net.URLClassLoader;
import java.nio.file.DirectoryStream;
import java.nio.file.DirectoryStream.Filter;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.util.Collection;
import java.util.HashSet;
import java.util.Set;
import java.util.jar.Attributes.Name;
import java.util.jar.JarEntry;
import java.util.jar.JarInputStream;

在包test中定义要扫描的类

package test;


public class A {
private class B {}


enum C {}


record D() {}
}

对于org.reflections:reflections:0.10.2,它为我工作如下:

使用反射库扫描包test中的类

@Test
void t() {
final String packagePath = "test";
final Reflections reflections =
new Reflections(packagePath, Scanners.SubTypes.filterResultsBy(v -> true));
reflections.getAll(Scanners.SubTypes).forEach(System.out::println);
}

输出

java.lang.constant.Constable
java.lang.Enum
java.lang.Comparable
java.lang.Record
java.lang.Object
java.io.Serializable
test.A$C
test.A$D
test.A$B
test.A

对于io.github.classgraph:classgraph:4.8.146,它为我工作如下:


@Test
void t() {
final String packagePath = "test";
try (ScanResult scanResult = new ClassGraph()
.enableClassInfo()
.ignoreClassVisibility()
.acceptPackages(packagePath)
.scan()) {
    

scanResult.getAllClasses()
.forEach(v -> {
System.out.println(v.getName());
});
}
}

输出

test.A
test.A$B
test.A$C
test.A$D