您可以使用反射找到包中的所有类吗？

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虚拟机它可以提供哪些类，它们只是对类的请求，并且必须返回一个类或抛出一个异常。

但是，如果您编写自己的类加载器，或检查类路径及其 jar，则可以找到此信息。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。

如果有生成或远程交付的类，您将无法发现这些类。

正常的方法是在某处注册您需要在文件中访问的类，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：The Reflections Library 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);

---

您可能应该看看开源 Reflections library。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);

---

Google Guava 14 包含一个新类 ClassPath，它具有三种扫描顶级类的方法：

getTopLevelClasses()

getTopLevelClasses（字符串包名）

getTopLevelClassesRecursive（字符串包名）

有关详细信息，请参阅 ClassPath javadocs。

---

您可以使用使用 ClassLoader 的 this method¹。

/**
 * 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，它是 Internet 档案馆的 archived，链接到现在。该代码段也可在 https://dzone.com/articles/get-all-classes-within-package 获得。}

---

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

注意：在版本 14 中，API 仍然标记为@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
  }
}

---

你好。我一直对上述解决方案（以及其他网站）有一些问题。作为开发人员，我正在为 API 编写插件。 API 阻止使用任何外部库或第 3 方工具。该设置还包含 jar 或 zip 文件中的代码以及直接位于某些目录中的类文件。因此，我的代码必须能够围绕每个设置工作。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。

编码：

此代码扫描给定包中包含的所有类。它仅适用于当前 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;
}

这三种方法使您能够找到给定包中的所有类。你像这样使用它：

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

说明：

该方法首先获取当前的 ClassLoader。然后它获取包含所述包的所有资源并迭代这些 URL。然后它创建一个 URLConnection 并确定我们拥有什么类型的 URl。它可以是目录 (FileURLConnection)，也可以是 jar 或 zip 文件 (JarURLConnection) 中的目录。根据连接的类型，我们将调用两种不同的方法。

首先让我们看看如果它是 FileURLConnection 会发生什么。
它首先检查传递的文件是否存在并且是一个目录。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 Class 对象并将其放入 ArrayList。如果它不是类文件而是目录，我们只需迭代它并做同样的事情。所有其他案例/文件将被忽略。

如果 URLConnection 是 JarURLConnection，则将调用另一个私有帮助方法。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在包内，则将创建一个 Class 对象并将其存储在 ArrayList 中。

解析完所有资源后，它（main 方法）返回包含给定包中所有类的 ArrayList，当前 ClassLoader 知道这些类。

如果该过程在任何时候失败，将抛出一个 ClassNotFoundException，其中包含有关确切原因的详细信息。

---

目前列出给定包中所有类的最强大的机制是 ClassGraph，因为它处理 widest possible array of classpath specification mechanisms，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());
}

---

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;
    }
}

---

一般来说，类加载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 文件列表（参见getURLs）并一一打开以列出可用的类。这种方法称为类路径扫描，在 Scannotation 和 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 编写注释处理器，它将在编译时收集所有带注释的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，由于 Java 编译器会自动发现在类路径中找到的任何处理器，因此扫描是完全自动化的，因此不需要额外的设置。

---

那这个呢：

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，您可以重载该函数：

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

如果你需要测试它：

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);
    }
}

如果您的 IDE 没有导入助手：

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;

有用：

从您的 IDE

对于 JAR 文件

没有外部依赖

---

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”，它会打印出来：

 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my other answer

---

这是我的做法。我扫描所有子文件夹（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，您可以使用很少的 API，这就是我喜欢这样做的方式，遇到了我使用休眠核心的问题，并且必须找到用某个注释进行注释的类。

使这些成为自定义注释，您将使用它标记要选择的类。

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

@EntityToBeScanned 
public MyClass{

}

制作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() 方法来获取一组找到的类。

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>

---

如果您在 Spring-land，您可以使用 PathMatchingResourcePatternResolver;

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

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

---

您需要查找类路径中的每个类加载器条目：

    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());
    }
}

现在一旦你有了包中的所有类名，你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，但是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。

我将发布在此重复问题中找到的解决方案：How to get all classes names in a package?

answer was written by sp00m;我添加了一些更正以使其正常工作：

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，只需调用本示例中提到的 sp00n 的 find 方法：如果需要，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，里面有测试方法，你可以看看~

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();
    }
}

---

几乎所有答案都使用 Reflections 或从文件系统读取类文件。如果您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);
}

---

使用 Maven 时，Aleksander Blomskøld's solution 不适用于参数化测试 @RunWith(Parameterized.class)。测试命名正确，并且在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

已报告类似问题 here。

在我的例子中，@Parameters 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，没有找到使用 Aleksander Blomskøld 解决方案的类。

我确实使它与以下片段一起工作，该片段的灵感来自 David Pärsson 对 Aleksander Blomskøld 的回答的评论：

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);

---

对于如此简单的事情，我找不到简短的工作片段。所以就在这里，我在搞砸了一段时间后自己做了：

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。

---

如果您没有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。

---

值得一提

如果您想拥有某个包下所有类的列表，可以使用 Reflection 如下方式：

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。

---

这是很有可能的，但是如果没有像 Reflections 这样的额外库，这很难...
这很难，因为您没有获取类名的完整工具。
而且，我采用了我的 {2 } 班级：

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;
}
}

---

这会扫描类加载器和所有父加载器以查找 jar 文件和目录。 jar 的 Class-Path 引用的 jar 文件和目录也被加载。

此代码是 Java 8,11,18 的 testet。在 8 上，使用 URLClassLoader 和 getURLs() 方法可以完美运行。在 11 上，它使用反射可以正常工作，但是 JVM 在 stderr 流上打印一个警告（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（在我知道它有效）是使用 getResource() 方法。当类加载器从文件系统加载给定包的资源时，会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;

---

在包测试中定义要扫描的类

package test;

public class A {
  private class B {}

  enum C {}

  record D() {}
}

对于 org.reflections:reflections:0.10.2，它适用于我如下：

使用反射库扫描包测试中的类

@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，它适用于我如下：

@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

---

基于 @Staale's answer，并且为了不依赖第三方库，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);

---

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 vavr 和 lombok 库

其他实现可以在 my GitHub daggerok/java-reflection-find-annotated-classes-or-methods repo 中找到

---

从 org.reflections 版本 0.10 开始：

org.reflections.scanners.SubTypesScanner 

和

org.reflections.Reflections.getAllTypes() 

已弃用。我用户：

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());
}

---

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。

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）所有您想要的类，而不管它们在什么包中。

另一方面，如果您想要将它们全部放在一个包中，那么只需让该包中的所有类都实现给定的接口。

请注意，接口本身不必声明任何方法 - 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。如果有兴趣了解如何阅读文档。

最后，我将提供另一种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来构建应用它的类的目录 - 类似于@ClassCatalog。

---

这是不可能的，因为可能不会加载包中的所有类，而您总是知道类的包。

---