dom4j 自帶一套解析 XML 文件的能力。本文用同一個解析器走完整條路:從零建立一份 XML 文件,再把它打開、修改,然後寫回磁碟。
清單 1 是我們要產出的原始文件 catalog.XML;清單 2 則是修改之後的 catalog-modified.XML。兩者結構相同,只有內容被替換過。
清單 1. 原始 XML 文件(catalog.XML)
<?xml version="1.0" encoding="UTF-8"?>
<catalog>
<!--An XML Catalog-->
<?target instruction?>
<journal title="XML Zone" publisher="IBM developerWorks">
<article level="Intermediate" date="December-2001">
<title>Java configuration with XML Schema</title>
<author>
<firstname>Marcello</firstname>
<lastname>Vitaletti</lastname>
</author>
</article>
</journal>
</catalog>
清單 2. 修改後的 XML 文件(catalog-modified.XML)
<?xml version="1.0" encoding="UTF-8"?>
<catalog>
<!--An XML Catalog-->
<?target instruction?>
<journal title="XML Zone" publisher="IBM developerWorks">
<article level="Introductory" date="October-2002">
<title>Create flexible and extensible XML schemas</title>
<author>
<firstname>Ayesha</firstname>
<lastname>Malik</lastname>
</author>
</article>
</journal>
</catalog>
dom4j 附帶的解析器有一個很實用的優勢:原生支援 XPath。W3C 的 DOM 解析器不提供以 XPath 挑節點的能力,文件結構稍深,取節點的程式碼就會迅速膨脹。
解析器可以從 https://dom4j.org 取得。最省事的做法是把 dom4j-1.4/dom4j-full.jar 放進 classpath,這支 jar 同時包含 dom4j 的類別、XPath 引擎,以及 SAX 與 DOM 介面。如果你的環境已經透過 JAXP 解析器提供了 SAX 與 DOM 介面,那麼只需要 dom4j-1.4/dom4j.jar——它同樣含 dom4j 類別與 XPath 引擎,但不打包 SAX 與 DOM 介面,可以避免版本衝突。
先 import 三個核心類別:
import org.dom4j.Document; import org.dom4j.DocumentHelper; import org.dom4j.Element;
接著用 DocumentHelper 這個工廠類別取得一個文件實例:
Document document = DocumentHelper.createDocument();
根元素 catalog 透過 addElement() 建立,這個方法同時也是往文件裡追加子元素的通用手段:
Element catalogElement = document.addElement("catalog");
註解與處理指令各有專屬方法:
catalogElement.addComment("An XML catalog");
catalogElement.addProcessingInstruction("target", "text");
往下一層,journal 元素掛在 catalog 之下,title 與 publisher 則是它的屬性,用 addAttribute() 加上:
Element journalElement = catalogElement.addElement("journal");
journalElement.addAttribute("title", "XML Zone");
journalElement.addAttribute("publisher", "IBM developerWorks");
article 元素同樣用 addElement() 接上,並補兩個屬性:
Element articleElement = journalElement.addElement("article");
articleElement.addAttribute("level", "Intermediate");
articleElement.addAttribute("date", "December-2001");
文字內容交給 setText():
Element titleElement = articleElement.addElement("title");
titleElement.setText("Java configuration with XML Schema");
作者資訊用一層 author 包住 firstname 與 lastname:
Element authorElement = articleElement.addElement("author");
Element firstNameElement = authorElement.addElement("firstname");
firstNameElement.setText("Marcello");
Element lastNameElement = authorElement.addElement("lastname");
lastNameElement.setText("Vitaletti");
如果文件需要以 DTD 驗證,別忘了加上文件型別宣告:
document.addDocType("catalog", null, "file://c:/Dtds/catalog.dtd");
它會產生這一行:
<!DOCTYPE catalog SYSTEM "file://c:/Dtds/catalog.dtd">
至於 <?xml version="1.0" encoding="UTF-8"?> 這行宣告,dom4j 會自動補上,不必手動處理。
清單 3. 產生 catalog.XML 的程式(XMLDom4J.java)
import org.dom4j.Document;
import org.dom4j.DocumentHelper;
import org.dom4j.Element;
import org.dom4j.io.XMLWriter;
import java.io.*;
public class XMLDom4J {
public void generateDocument() {
Document document = DocumentHelper.createDocument();
Element catalogElement = document.addElement("catalog");
catalogElement.addComment("An XML Catalog");
catalogElement.addProcessingInstruction("target", "text");
Element journalElement = catalogElement.addElement("journal");
journalElement.addAttribute("title", "XML Zone");
journalElement.addAttribute("publisher", "IBM developerWorks");
Element articleElement = journalElement.addElement("article");
articleElement.addAttribute("level", "Intermediate");
articleElement.addAttribute("date", "December-2001");
Element titleElement = articleElement.addElement("title");
titleElement.setText("Java configuration with XML Schema");
Element authorElement = articleElement.addElement("author");
Element firstNameElement = authorElement.addElement("firstname");
firstNameElement.setText("Marcello");
Element lastNameElement = authorElement.addElement("lastname");
lastNameElement.setText("Vitaletti");
document.addDocType("catalog", null, "file://c:/Dtds/catalog.dtd");
try {
XMLWriter output = new XMLWriter(
new FileWriter(new File("c:/catalog/catalog.XML")));
output.write(document);
output.close();
} catch (IOException e) {
System.out.println(e.getMessage());
}
}
public static void main(String[] argv) {
XMLDom4J dom4j = new XMLDom4J();
dom4j.generateDocument();
}
}
修改的起點是把文件讀進來。SAXReader 位於 org.dom4j.io 套件,read() 接受一個 File:
SAXReader saxReader = new SAXReader(); Document document = saxReader.read(inputXML);
這裡的 inputXml 指向 c:/catalog/catalog.XML。接下來示範幾種常見的修改動作。
用 XPath 取出所有 article 的 level 屬性,把 Intermediate 換成 Introductory:
List list = document.selectNodes("//article/@level");
Iterator iter = list.iterator();
while (iter.hasNext()) {
Attribute attribute = (Attribute) iter.next();
if (attribute.getValue().equals("Intermediate"))
attribute.setValue("Introductory");
}
先取 article 清單,再用 elementIterator("title") 逐一檢視它的 title 子元素:
list = document.selectNodes("//article");
iter = list.iterator();
while (iter.hasNext()) {
Element element = (Element) iter.next();
Iterator iterator = element.elementIterator("title");
while (iterator.hasNext()) {
Element titleElement = (Element) iterator.next();
if (titleElement.getText().equals("Java configuration with XML Schema"))
titleElement.setText("Create flexible and extensible XML schema");
}
}
author 底下的 firstname 與 lastname,用完全相同的套路處理即可。
清單 4. 把 catalog.XML 改成 catalog-modified.XML 的程式(Dom4JParser.java)
import org.dom4j.Attribute;
import org.dom4j.Document;
import org.dom4j.DocumentException;
import org.dom4j.Element;
import org.dom4j.io.SAXReader;
import org.dom4j.io.XMLWriter;
import java.io.*;
import java.util.Iterator;
import java.util.List;
public class Dom4JParser {
public void modifyDocument(File inputXML) {
try {
SAXReader saxReader = new SAXReader();
Document document = saxReader.read(inputXML);
// 1) 改屬性:level
List list = document.selectNodes("//article/@level");
Iterator iter = list.iterator();
while (iter.hasNext()) {
Attribute attribute = (Attribute) iter.next();
if (attribute.getValue().equals("Intermediate"))
attribute.setValue("Introductory");
}
// 2) 改屬性:date
list = document.selectNodes("//article/@date");
iter = list.iterator();
while (iter.hasNext()) {
Attribute attribute = (Attribute) iter.next();
if (attribute.getValue().equals("December-2001"))
attribute.setValue("October-2002");
}
// 3) 改元素文字:title
list = document.selectNodes("//article");
iter = list.iterator();
while (iter.hasNext()) {
Element element = (Element) iter.next();
Iterator iterator = element.elementIterator("title");
while (iterator.hasNext()) {
Element titleElement = (Element) iterator.next();
if (titleElement.getText()
.equals("Java configuration with XML Schema"))
titleElement.setText(
"Create flexible and extensible XML schema");
}
}
// 4) 改元素文字:author/firstname
list = document.selectNodes("//article/author");
iter = list.iterator();
while (iter.hasNext()) {
Element element = (Element) iter.next();
Iterator iterator = element.elementIterator("firstname");
while (iterator.hasNext()) {
Element firstNameElement = (Element) iterator.next();
if (firstNameElement.getText().equals("Marcello"))
firstNameElement.setText("Ayesha");
}
}
// 5) 改元素文字:author/lastname
list = document.selectNodes("//article/author");
iter = list.iterator();
while (iter.hasNext()) {
Element element = (Element) iter.next();
Iterator iterator = element.elementIterator("lastname");
while (iterator.hasNext()) {
Element lastNameElement = (Element) iterator.next();
if (lastNameElement.getText().equals("Vitaletti"))
lastNameElement.setText("Malik");
}
}
XMLWriter output = new XMLWriter(
new FileWriter(new File("c:/catalog/catalog-modified.XML")));
output.write(document);
output.close();
} catch (DocumentException e) {
System.out.println(e.getMessage());
} catch (IOException e) {
System.out.println(e.getMessage());
}
}
public static void main(String[] argv) {
Dom4JParser dom4jParser = new Dom4JParser();
dom4jParser.modifyDocument(new File("c:/catalog/catalog.XML"));
}
}
要留意的是,這個解析器不做 DTD 或結構描述驗證。若你的流程需要驗證,可以搭配 dom4j 與 JAXP 的 SAX 解析器一起使用。
dom4j 內附的是一套非驗證式的 XML 解析工具,但由於它能與 JAXP、Crimson 或 Xerces 等既有實作整合,實際專案裡很少需要為了驗證而把它換掉。把建立、讀取、修改這三件事收在同一個 API 之下,再加上原生 XPath 的加持,處理 XML 的程式碼會比想像中清爽得多。