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());
}

在提出我自己的解决方案之前，我应该先看看这一页!不管怎样，我使用Java递归来解析xml页面。此代码是完全自包含的，不依赖于第三方库。也. .它使用递归!

// you call this method passing in the xml text
public static void prettyPrint(String text){
    prettyPrint(text, 0);
}

// "index" corresponds to the number of levels of nesting and/or the number of tabs to print before printing the tag
public static void prettyPrint(String xmlText, int index){
    boolean foundTagStart = false;
    StringBuilder tagChars = new StringBuilder();
    String startTag = "";
    String endTag = "";
    String[] chars = xmlText.split("");
    // find the next start tag
    for(String ch : chars){
        if(ch.equalsIgnoreCase("<")){
            tagChars.append(ch);
            foundTagStart = true;
        } else if(ch.equalsIgnoreCase(">") && foundTagStart){
            startTag = tagChars.append(ch).toString();
            String tempTag = startTag;
            endTag = (tempTag.contains("\"") ? (tempTag.split(" ")[0] + ">") : tempTag).replace("<", "</"); // <startTag attr1=1 attr2=2> => </startTag>
            break;
        } else if(foundTagStart){
            tagChars.append(ch);
        }
    }
    // once start and end tag are calculated, print start tag, then content, then end tag
    if(foundTagStart){
        int startIndex = xmlText.indexOf(startTag);
        int endIndex = xmlText.indexOf(endTag);
        // handle if matching tags NOT found
        if((startIndex < 0) || (endIndex < 0)){
            if(startIndex < 0) {
                // no start tag found
                return;
            } else {
                // start tag found, no end tag found (handles single tags aka "<mytag/>" or "<?xml ...>")
                printTabs(index);
                System.out.println(startTag);
                // move on to the next tag
                // NOTE: "index" (not index+1) because next tag is on same level as this one
                prettyPrint(xmlText.substring(startIndex+startTag.length(), xmlText.length()), index);
                return;
            }
        // handle when matching tags found
        } else {
            String content = xmlText.substring(startIndex+startTag.length(), endIndex);
            boolean isTagContainsTags = content.contains("<"); // content contains tags
            printTabs(index);
            if(isTagContainsTags){ // ie: <tag1><tag2>stuff</tag2></tag1>
                System.out.println(startTag);
                prettyPrint(content, index+1); // "index+1" because "content" is nested
                printTabs(index);
            } else {
                System.out.print(startTag); // ie: <tag1>stuff</tag1> or <tag1></tag1>
                System.out.print(content);
            }
            System.out.println(endTag);
            int nextIndex = endIndex + endTag.length();
            if(xmlText.length() > nextIndex){ // if there are more tags on this level, continue
                prettyPrint(xmlText.substring(nextIndex, xmlText.length()), index);
            }
        }
    } else {
        System.out.print(xmlText);
    }
}

private static void printTabs(int counter){
    while(counter-- > 0){ 
        System.out.print("\t");
    }
}

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);
}

我过去使用org.dom4j.io.OutputFormat.createPrettyPrint()方法打印过

public String prettyPrint(final String xml){  

    if (StringUtils.isBlank(xml)) {
        throw new RuntimeException("xml was null or blank in prettyPrint()");
    }

    final StringWriter sw;

    try {
        final OutputFormat format = OutputFormat.createPrettyPrint();
        final org.dom4j.Document document = DocumentHelper.parseText(xml);
        sw = new StringWriter();
        final XMLWriter writer = new XMLWriter(sw, format);
        writer.write(document);
    }
    catch (Exception e) {
        throw new RuntimeException("Error pretty printing xml:\n" + xml, e);
    }
    return sw.toString();
}

稍微改进了milosmns的版本…

public static String getPrettyXml(String xml) {
    if (xml == null || xml.trim().length() == 0) return "";

    int stack = 0;
    StringBuilder pretty = new StringBuilder();
    String[] rows = xml.trim().replaceAll(">", ">\n").replaceAll("<", "\n<").split("\n");

    for (int i = 0; i < rows.length; i++) {
        if (rows[i] == null || rows[i].trim().length() == 0) continue;

        String row = rows[i].trim();
        if (row.startsWith("<?")) {
            pretty.append(row + "\n");
        } else if (row.startsWith("</")) {
            String indent = repeatString(--stack);
            pretty.append(indent + row + "\n");
        } else if (row.startsWith("<") && row.endsWith("/>") == false) {
            String indent = repeatString(stack++);
            pretty.append(indent + row + "\n");
            if (row.endsWith("]]>")) stack--;
        } else {
            String indent = repeatString(stack);
            pretty.append(indent + row + "\n");
        }
    }

    return pretty.toString().trim();
}

private static String repeatString(int stack) {
     StringBuilder indent = new StringBuilder();
     for (int i = 0; i < stack; i++) {
        indent.append(" ");
     }
     return indent.toString();
} 

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());