为了将来的参考，这里有一个对我有用的解决方案(感谢@George Hawkins在其中一个答案中发表的评论):

DOMImplementationRegistry registry = DOMImplementationRegistry.newInstance();
DOMImplementationLS impl = (DOMImplementationLS) registry.getDOMImplementation("LS");
LSSerializer writer = impl.createLSSerializer();
writer.getDomConfig().setParameter("format-pretty-print", Boolean.TRUE);
LSOutput output = impl.createLSOutput();
ByteArrayOutputStream out = new ByteArrayOutputStream();
output.setByteStream(out);
writer.write(document, output);
String xmlStr = new String(out.toByteArray());

我在这里找到的针对Java 1.6+的解决方案不会重新格式化已经格式化的代码。对我有效的方法(重新格式化已经格式化的代码)如下。

import org.apache.xml.security.c14n.CanonicalizationException;
import org.apache.xml.security.c14n.Canonicalizer;
import org.apache.xml.security.c14n.InvalidCanonicalizerException;
import org.w3c.dom.Element;
import org.w3c.dom.bootstrap.DOMImplementationRegistry;
import org.w3c.dom.ls.DOMImplementationLS;
import org.w3c.dom.ls.LSSerializer;
import org.xml.sax.InputSource;
import org.xml.sax.SAXException;

import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.parsers.ParserConfigurationException;
import javax.xml.transform.TransformerException;
import java.io.IOException;
import java.io.StringReader;

public class XmlUtils {
    public static String toCanonicalXml(String xml) throws InvalidCanonicalizerException, ParserConfigurationException, SAXException, CanonicalizationException, IOException {
        Canonicalizer canon = Canonicalizer.getInstance(Canonicalizer.ALGO_ID_C14N_OMIT_COMMENTS);
        byte canonXmlBytes[] = canon.canonicalize(xml.getBytes());
        return new String(canonXmlBytes);
    }

    public static String prettyFormat(String input) throws TransformerException, ParserConfigurationException, IOException, SAXException, InstantiationException, IllegalAccessException, ClassNotFoundException {
        InputSource src = new InputSource(new StringReader(input));
        Element document = DocumentBuilderFactory.newInstance().newDocumentBuilder().parse(src).getDocumentElement();
        Boolean keepDeclaration = input.startsWith("<?xml");
        DOMImplementationRegistry registry = DOMImplementationRegistry.newInstance();
        DOMImplementationLS impl = (DOMImplementationLS) registry.getDOMImplementation("LS");
        LSSerializer writer = impl.createLSSerializer();
        writer.getDomConfig().setParameter("format-pretty-print", Boolean.TRUE);
        writer.getDomConfig().setParameter("xml-declaration", keepDeclaration);
        return writer.writeToString(document);
    }
}

它是在单元测试中比较全字符串xml的好工具。

private void assertXMLEqual(String expected, String actual) throws ParserConfigurationException, IOException, SAXException, CanonicalizationException, InvalidCanonicalizerException, TransformerException, IllegalAccessException, ClassNotFoundException, InstantiationException {
    String canonicalExpected = prettyFormat(toCanonicalXml(expected));
    String canonicalActual = prettyFormat(toCanonicalXml(actual));
    assertEquals(canonicalExpected, canonicalActual);
}

关于“您必须首先构建DOM树”的评论:不，您不需要也不应该这样做。

相反，创建一个StreamSource(new StreamSource(new StringReader(str))，并将其提供给前面提到的标识转换器。这将使用SAX解析器，结果将快得多。 在这种情况下，构建中间树纯粹是开销。 否则，排名第一的答案是好的。

我试图实现类似的东西，但没有任何外部依赖。应用程序已经在使用DOM来格式化xml了!

下面是我的示例片段

public void formatXML(final String unformattedXML) {
    final int length = unformattedXML.length();
    final int indentSpace = 3;
    final StringBuilder newString = new StringBuilder(length + length / 10);
    final char space = ' ';
    int i = 0;
    int indentCount = 0;
    char currentChar = unformattedXML.charAt(i++);
    char previousChar = currentChar;
    boolean nodeStarted = true;
    newString.append(currentChar);
    for (; i < length - 1;) {
        currentChar = unformattedXML.charAt(i++);
        if(((int) currentChar < 33) && !nodeStarted) {
            continue;
        }
        switch (currentChar) {
        case '<':
            if ('>' == previousChar && '/' != unformattedXML.charAt(i - 1) && '/' != unformattedXML.charAt(i) && '!' != unformattedXML.charAt(i)) {
                indentCount++;
            }
            newString.append(System.lineSeparator());
            for (int j = indentCount * indentSpace; j > 0; j--) {
                newString.append(space);
            }
            newString.append(currentChar);
            nodeStarted = true;
            break;
        case '>':
            newString.append(currentChar);
            nodeStarted = false;
            break;
        case '/':
            if ('<' == previousChar || '>' == unformattedXML.charAt(i)) {
                indentCount--;
            }
            newString.append(currentChar);
            break;
        default:
            newString.append(currentChar);
        }
        previousChar = currentChar;
    }
    newString.append(unformattedXML.charAt(length - 1));
    System.out.println(newString.toString());
}

下面的代码工作得很好

import javax.xml.transform.OutputKeys;
import javax.xml.transform.Source;
import javax.xml.transform.Transformer;
import javax.xml.transform.TransformerFactory;
import javax.xml.transform.stream.StreamResult;
import javax.xml.transform.stream.StreamSource;

String formattedXml1 = prettyFormat("<root><child>aaa</child><child/></root>");

public static String prettyFormat(String input) {
    return prettyFormat(input, "2");
}

public static String prettyFormat(String input, String indent) {
    Source xmlInput = new StreamSource(new StringReader(input));
    StringWriter stringWriter = new StringWriter();
    try {
        TransformerFactory transformerFactory = TransformerFactory.newInstance();
        Transformer transformer = transformerFactory.newTransformer();
        transformer.setOutputProperty(OutputKeys.INDENT, "yes");
        transformer.setOutputProperty("{http://xml.apache.org/xslt}indent-amount", indent);
        transformer.transform(xmlInput, new StreamResult(stringWriter));

        String pretty = stringWriter.toString();
        pretty = pretty.replace("\r\n", "\n");
        return pretty;              
    } catch (Exception e) {
        throw new RuntimeException(e);
    }
}

Since you are starting with a String, you can convert to a DOM object (e.g. Node) before you use the Transformer. However, if you know your XML string is valid, and you don't want to incur the memory overhead of parsing a string into a DOM, then running a transform over the DOM to get a string back - you could just do some old fashioned character by character parsing. Insert a newline and spaces after every </...> characters, keep and indent counter (to determine the number of spaces) that you increment for every <...> and decrement for every </...> you see.

免责声明-我对下面的函数做了剪切/粘贴/文本编辑，所以它们可能不能按原样编译。

public static final Element createDOM(String strXML) 
    throws ParserConfigurationException, SAXException, IOException {

    DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();
    dbf.setValidating(true);
    DocumentBuilder db = dbf.newDocumentBuilder();
    InputSource sourceXML = new InputSource(new StringReader(strXML));
    Document xmlDoc = db.parse(sourceXML);
    Element e = xmlDoc.getDocumentElement();
    e.normalize();
    return e;
}

public static final void prettyPrint(Node xml, OutputStream out)
    throws TransformerConfigurationException, TransformerFactoryConfigurationError, TransformerException {
    Transformer tf = TransformerFactory.newInstance().newTransformer();
    tf.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "yes");
    tf.setOutputProperty(OutputKeys.ENCODING, "UTF-8");
    tf.setOutputProperty(OutputKeys.INDENT, "yes");
    tf.transform(new DOMSource(xml), new StreamResult(out));
}