你能使用反射在一个包中找到所有类吗？

这是不可能的，因为包中的所有类可能不会被加载，而你总是知道一个类的包。

由于类加载器的动态特性，这是不可能的。类加载器不需要告诉 VM 它可以提供哪些类，而只是向类发出请求，并且必须返回类或引发异常。

但是，如果您编写自己的类加载器，或者检查类路径及其 jar，则可以找到此信息。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您做到这一点。

如果存在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您需要访问的类，或在不同的类中引用它们。或者只是在命名时使用约定。

附录：反射库将允许您在当前类路径中查找类。它可用于获取包中的所有类：

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

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

您可以使用此方法 ¹，该方法使用 .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;
}

__________
_{¹ 此方法最初取自 http://snippets.dzone.com/posts/show/4831，该 已由互联网档案馆存档，链接到现在。该代码片段也可在 https://dzone.com/articles/get-all-classes-within-package 上获得。}

如果您没有使用任何动态类装入器，则可以搜索类路径，并为每个条目搜索目录或 JAR 文件。

您可能应该看看开源的 Reflections 库。有了它，您可以轻松实现您想要的。

首先，设置反射索引（这有点混乱，因为默认情况下禁用搜索所有类）：

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，您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

它应该适用于基于文件的类路径和 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. 

另请参阅我的其他答案

Google Guava 14 包含一个新类 ClassPath，其中包含三种扫描顶级类的方法：

• getTopLevelClasses()
• getTopLevelClasses(String packageName)
• getTopLevelClassesRecursive(String packageName)

有关详细信息，请参阅 ClassPath javadocs。

通常，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 and Reflections 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 Pluggable Annotation Processing API 来编写注解处理器，该处理器将在编译时收集所有注解类并构建索引文件以供运行时使用。此机制在 ClassIndex 库中实现：

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

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

请注意，不需要额外的设置，因为扫描是完全自动化的，这要归功于 Java 编译器自动发现在类路径上找到的任何处理器。

您需要在类路径中查找每个类装入器条目：

    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，只需在正确的子目录中查找：

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，并且是 jar，请检查它的 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，一旦你在包中拥有了所有类名，你就可以尝试用反射加载它们，并分析它们是类还是接口等。

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

我将发布我在这个重复的问题中找到的解决方案：如何获取包中的所有类名？

答案是由 sp00m 编写的;我添加了一些更正以使其工作：

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;
    }

}

若要使用它，只需调用此示例中提到的 sp00n 的 find 方法： 如果需要，我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，里面有测试方法，可以有查一下~

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

此示例适用于 Spring 4，但您也可以在早期版本中找到类路径扫描程序。

// 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

注意：在版本 14 中，API 仍标记为@Beta，因此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
  }
}

在不使用任何额外库的情况下：

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。我总是对上述解决方案（以及其他网站）有一些问题。
作为一名开发人员，我正在为 API 编写一个插件。该 API 阻止使用任何外部库或第三方工具。该设置还包含 jar 或 zip 文件中的代码和直接位于某些目录中的类文件的混合。因此，我的代码必须能够适应每个设置。经过大量研究，我想出了一种方法，该方法至少适用于所有可能设置的 95%。

下面的代码基本上是永远有效的矫枉过正的方法。

代码：

此代码扫描给定包中包含的所有类。它仅适用于当前 .ClassLoader

/**
 * 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;
}

这三种方法使您能够在给定的包中查找所有类。
你像这样使用它：

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

解释：

该方法首先获取当前 .然后，它获取包含所述包的所有资源并迭代这些包。然后，它创建一个并确定我们拥有的 URl 类型。它可以是目录 （），也可以是 jar 或 zip 文件 （） 中的目录。根据我们拥有的连接类型，将调用两种不同的方法。ClassLoaderURLURLConnectionFileURLConnectionJarURLConnection

首先，让我们看看如果它是 .
它首先检查传递的文件是否存在并且是否为目录。如果是这种情况，它会检查它是否是类文件。如果是这样，将创建一个对象并将其放入 .如果它不是类文件而是目录，我们只需迭代它并执行相同的操作。所有其他案例/文件将被忽略。FileURLConnectionClassArrayList

如果 是 ，则将调用其他私人帮助程序方法。此方法遍历 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，则将创建一个对象并将其存储在 .URLConnectionJarURLConnectionClassArrayList

解析完所有资源后，它（main 方法）返回当前知道的给定包中的所有类。ArrayListClassLoader

如果该过程在任何时候失败，将抛出包含有关确切原因的详细信息。ClassNotFoundException

几乎所有的答案都使用或从文件系统中读取类文件。如果尝试从文件系统中读取类，则在将应用程序打包为 JAR 或其他应用程序时可能会出现错误。此外，您可能不希望为此目的使用单独的库。Reflections

这是另一种纯 java 方法，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));
    }
}

Java 8 不是必须的。您可以使用 for 循环而不是流。 你可以像这样测试它

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);
}

^{值得一提的是}

如果你想在某个包下有一个所有类的列表，你可以使用以下方式：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，稍后您可以根据需要使用它们。

目前，列出给定包中所有类的最健壮的机制是 ClassGraph，因为它可以处理尽可能广泛的类路径规范机制阵列，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());
}

这是我的做法。我扫描了所有子文件夹（子包），并且不尝试加载匿名类：

   /**
   * 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，这就是我喜欢的做法，面对这个问题，我正在使用休眠核心，并且必须找到用某种注释进行注释的类。

将这些作为自定义注释，您将使用它来标记要选取的类。

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 {

}

然后用它标记你的班级，比如

@EntityToBeScanned 
public MyClass{

}

使这个实用程序类具有以下方法

public class ClassScanner {

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

}

调用 allFoundClassesAnnotatedWithEntityToBeScanned（） 方法以获取找到的一组类。

您将需要下面给出的库

<!-- 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>

这是很有可能的，但是如果没有像这样的额外库，这很难......
这很难，因为你没有完整的工具来获取类名。
而且，我采用我的类代码：ReflectionsClassFinder

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;
}
}

Aleksander Blomskøld 的解决方案在使用 Maven 时不适用于参数化测试。测试的名称正确，以及找到但未执行的地方：@RunWith(Parameterized.class)

-------------------------------------------------------
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

此处报告了类似的问题。

就我而言，是在包中创建每个类的实例。在 IDE 中本地运行时，测试运行良好。但是，在运行 Maven 时，没有发现 Aleksander Blomskøld 解决方案的类。@Parameters

我确实用以下片段让它工作，这是受 David Pärsson 对 Aleksander Blomskøld 答案的评论的启发：

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 可以为您提供帮助。

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

话虽如此，作为从文件系统中加载类列表的替代方法，只需在要加载的所有类中实现相同的接口，而不管包如何，并使用 Spring 为您提供所有这些类的实例。这样，您可以加载（和实例化）所需的所有类，无论它们在哪个包中。

另一方面，如果将它们全部放在一个包中是你想要的，那么只需让该包中的所有类实现一个给定的接口。

请注意，接口本身不必声明任何方法 - 它可以是完全空的。

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，了解如何操作。

最后，我将提供另一种解决方案，它比搜索整个文件系统树更优雅一些。

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

纯 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：为了简化处理错误等的样板，我在这里使用和库vavrlombok

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

这个怎么样：

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，您可以重载该函数：

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

如果需要测试它：

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 IDE 没有导入帮助程序：

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;

它的工作原理：

• 从 IDE

• 对于 JAR 文件

• 无外部依赖

对于如此简单的事情，我找不到一个简短的工作片段。所以这里是我自己搞砸了一段时间后制作的：

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->{
        ...
    });

截至版本：org.reflections0.10

org.reflections.scanners.SubTypesScanner 

和

org.reflections.Reflections.getAllTypes() 

已弃用。我使用：

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());
}

这将扫描类加载器和所有父加载器以查找 JAR 文件和目录。 还会加载 jar 的 Class-Path 引用的 jar 文件和目录。

此代码是 Java 8、11、18 的测试版。 在 8 上，使用 URLClassLoader 和 getURLs（） 方法，一切都可以完美运行。 在 11 上，它使用反射工作正常，但 JVM 在 stderr 流上打印警告（不能用我的 JVM 使用 System.setErr（） 重定向） 在 18 上，反射是无用的（抛出 NoSuchMethod/Field），唯一的事情（我知道它有效的地方）是使用 getResource（） 方法。当类装入器从文件系统加载给定包的资源时，将返回一个简单的路径 url。当类加载器从 jar 加载资源时，会返回类似 'jar：file：[jar-path]！/[in-jar-path]' 的 url。

我使用了答案 https://stackoverflow.com/a/1157352/18252455（来自重复的问题）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;

定义要在包中扫描的类test

package test;

public class A {
  private class B {}

  enum C {}

  record D() {}
}

对于 ，它对我的工作方式如下：org.reflections:reflections:0.10.2

使用 reflection lib 扫描包中的类test

@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 ，它对我的工作方式如下：io.github.classgraph:classgraph:4.8.146

@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