org.argeo.cms/src/org/argeo/cms/acr/xml/XmlNormalizer.java

   1 package org.argeo.cms.acr.xml;
   2
   3 import java.io.ByteArrayInputStream;
   4 import java.io.IOException;
   5 import java.io.InputStream;
   6 import java.io.OutputStream;
   7 import java.nio.file.DirectoryStream;
   8 import java.nio.file.Files;
   9 import java.nio.file.Path;
  10 import java.nio.file.Paths;
  11
  12 import javax.xml.parsers.DocumentBuilder;
  13 import javax.xml.parsers.DocumentBuilderFactory;
  14 import javax.xml.parsers.ParserConfigurationException;
  15 import javax.xml.transform.OutputKeys;
  16 import javax.xml.transform.Transformer;
  17 import javax.xml.transform.TransformerConfigurationException;
  18 import javax.xml.transform.TransformerException;
  19 import javax.xml.transform.TransformerFactory;
  20 import javax.xml.transform.TransformerFactoryConfigurationError;
  21 import javax.xml.transform.dom.DOMSource;
  22 import javax.xml.transform.stream.StreamResult;
  23 import javax.xml.transform.stream.StreamSource;
  24
  25 import org.w3c.dom.DOMException;
  26 import org.w3c.dom.Document;
  27 import org.xml.sax.SAXException;
  28
  29 /**
  30  * Consistently normalises an XML in order to ease diff (typically in a
  31  * versioning system).
  32  */
  33 public class XmlNormalizer {
  34 //      private final static Logger logger = System.getLogger(XmlNormalizer.class.getName());
  35
  36         private DocumentBuilder documentBuilder;
  37         private Transformer transformer;
  38
  39         public XmlNormalizer() {
  40                 this(2);
  41         }
  42
  43         public XmlNormalizer(int indent) {
  44                 try {
  45                         documentBuilder = DocumentBuilderFactory.newNSInstance().newDocumentBuilder();
  46                         TransformerFactory transformerFactory = TransformerFactory.newInstance();
  47                         transformerFactory.setAttribute("indent-number", indent);
  48                         try (InputStream in = XmlNormalizer.class.getResourceAsStream("stripSpaces.xsl")) {
  49                                 transformer = transformerFactory.newTransformer(new StreamSource(in));
  50                         }
  51                 } catch (TransformerConfigurationException | ParserConfigurationException | TransformerFactoryConfigurationError
  52                                 | IOException e) {
  53                         throw new IllegalStateException("Cannot initialise document builder and transformer", e);
  54                 }
  55         }
  56
  57         public void normalizeXmlFiles(DirectoryStream<Path> ds) throws IOException {
  58                 for (Path path : ds) {
  59                         normalizeXmlFile(path);
  60                 }
  61         }
  62
  63         public void normalizeXmlFile(Path path) throws IOException {
  64                 byte[] bytes = Files.readAllBytes(path);
  65                 try (ByteArrayInputStream in = new ByteArrayInputStream(bytes);
  66                                 OutputStream out = Files.newOutputStream(path)) {
  67                         normalizeAndIndent(in, out);
  68 //                      logger.log(Level.DEBUG, () -> "Normalized XML " + path);
  69                 }
  70         }
  71
  72         public void normalizeAndIndent(InputStream in, OutputStream out) throws IOException {
  73                 normalizeAndIndent(in, out, 2);
  74         }
  75
  76         public void normalizeAndIndent(InputStream in, OutputStream out, int indent) throws IOException {
  77                 try {
  78                         Document document = documentBuilder.parse(in);
  79
  80                         // clear whitespaces outside tags
  81                         document.normalize();
  82 //                      XPath xPath = XPathFactory.newInstance().newXPath();
  83 //                      NodeList nodeList = (NodeList) xPath.evaluate("//text()[normalize-space()='']", document,
  84 //                                      XPathConstants.NODESET);
  85 //
  86 //                      for (int i = 0; i < nodeList.getLength(); ++i) {
  87 //                              Node node = nodeList.item(i);
  88 //                              node.getParentNode().removeChild(node);
  89 //                      }
  90
  91                         transformer.setOutputProperty(OutputKeys.ENCODING, "UTF-8");
  92                         // transformer.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "yes");
  93                         transformer.setOutputProperty(OutputKeys.INDENT, "yes");
  94
  95                         transformer.transform(new DOMSource(document), new StreamResult(out));
  96                 } catch (DOMException | IllegalArgumentException | SAXException | TransformerFactoryConfigurationError
  97                                 | TransformerException e) {
  98                         throw new RuntimeException("Cannot normalise and indent XML", e);
  99                 }
 100         }
 101
 102         public static void main(String[] args) throws IOException {
 103                 Path dir = Paths.get(args[0]);
 104                 XmlNormalizer xmlNormalizer = new XmlNormalizer();
 105                 DirectoryStream<Path> ds = Files.newDirectoryStream(dir, "*.svg");
 106                 xmlNormalizer.normalizeXmlFiles(ds);
 107
 108         }
 109 }