Java 类org.jsoup.nodes.Document.OutputSettings 实例源码

项目：crawler-jsoup-maven 文件：JsoupTest.java

public static void main(String[] args) {

    String d = "<span><div>test</div></span>";
    Document doc = Jsoup.parse(d);
    Element div = doc.select("div").first(); // <div></div>
    div.html("<p>lorem ipsum</p>"); // <div><p>lorem ipsum</p></div>
    div.prepend("<p>First</p>");
    div.append("<p>Last</p>");
    // now: <div><p>First</p><p>lorem ipsum</p><p>Last</p></div>
    div.appendElement(d);
    Element span = doc.select("span").first(); // <span>One</span>
    span.wrap("<li><a href='http://example.com/'></a></li>");
    // now: <li><a href="http://example.com"><span>One</span></a></li>
    System.out.println(doc.html());

    String s = Jsoup.clean(doc.html(), "", Whitelist.relaxed(), new OutputSettings().prettyPrint(false));

    System.out.println(s);
}

项目：docx4j-template 文件：XHTMLDocumentHandler.java

/**
 * Jsoup.parse(in, charsetName, baseUri)
 */
@Override
public Document handle( InputStream input) throws IOException{
    //获取Jsoup参数
    String charsetName = Docx4jProperties.getProperty(Docx4jConstants.DOCX4J_JSOUP_PARSE_CHARSETNAME, Docx4jConstants.DEFAULT_CHARSETNAME );
    String baseUri = Docx4jProperties.getProperty(Docx4jConstants.DOCX4J_JSOUP_PARSE_BASEURI,"");
    //使用Jsoup将html转换成Document对象
    Document doc = Jsoup.parse(input, charsetName, baseUri);

    OutputSettings outputSettings = new OutputSettings();

    outputSettings.prettyPrint(false);

    /*
    outputSettings.syntax(syntax)
    outputSettings.charset(charset)
    outputSettings*/
    doc.outputSettings(outputSettings);

    //返回Document对象
    return doc;
}

项目：mesfavoris 文件：JavadocCommentProvider.java

private String getJavadocCommentAsText(IMember member) {
    try (Reader reader = JavadocContentAccess.getHTMLContentReader(member, true, true)) {
        if (reader == null) {
            return null;
        }
        String javadocAsHtml = CharStreams.toString(reader);
        String javadocAsString = Jsoup.clean(javadocAsHtml, "", Whitelist.none(), new OutputSettings().prettyPrint(false));

        // trim lines
        try (BufferedReader bufferedReader = new BufferedReader(new StringReader(javadocAsString))) {
            return bufferedReader.lines().map(line->line.trim()).collect(Collectors.joining("\n"));
        }
    } catch (JavaModelException | IOException e) {
        return null;
    }
}

项目：common 文件：DocumentTest.java

@Test public void testHtmlAndXmlSyntax() {
    String h = "<!DOCTYPE html><body><img async checked='checked' src='&<>\"'>&lt;&gt;&amp;&quot;<foo />bar";
    Document doc = Jsoup.parse(h);

    doc.outputSettings().syntax(Syntax.html);
    assertEquals("<!doctype html>\n" +
            "<html>\n" +
            " <head></head>\n" +
            " <body>\n" +
            "  <img async checked src=\"&amp;<>&quot;\">&lt;&gt;&amp;\"\n" +
            "  <foo />bar\n" +
            " </body>\n" +
            "</html>", doc.html());

    doc.outputSettings().syntax(Document.OutputSettings.Syntax.xml);
    assertEquals("<!DOCTYPE html>\n" +
            "<html>\n" +
            " <head></head>\n" +
            " <body>\n" +
            "  <img async=\"\" checked=\"checked\" src=\"&amp;<>&quot;\" />&lt;&gt;&amp;\"\n" +
            "  <foo />bar\n" +
            " </body>\n" +
            "</html>", doc.html());
}

项目：astor 文件：DocumentTest.java

@Test public void testHtmlAndXmlSyntax() {
    String h = "<!DOCTYPE html><body><img async checked='checked' src='&<>\"'>&lt;&gt;&amp;&quot;<foo />bar";
    Document doc = Jsoup.parse(h);

    doc.outputSettings().syntax(Syntax.html);
    assertEquals("<!doctype html>\n" +
            "<html>\n" +
            " <head></head>\n" +
            " <body>\n" +
            "  <img async checked src=\"&amp;<>&quot;\">&lt;&gt;&amp;\"\n" +
            "  <foo />bar\n" +
            " </body>\n" +
            "</html>", doc.html());

    doc.outputSettings().syntax(Document.OutputSettings.Syntax.xml);
    assertEquals("<!DOCTYPE html>\n" +
            "<html>\n" +
            " <head></head>\n" +
            " <body>\n" +
            "  <img async=\"\" checked=\"checked\" src=\"&amp;<>&quot;\" />&lt;&gt;&amp;\"\n" +
            "  <foo />bar\n" +
            " </body>\n" +
            "</html>", doc.html());
}

项目：astor 文件：DocumentTest.java

@Test public void testHtmlAndXmlSyntax() {
    String h = "<!DOCTYPE html><body><img async checked='checked' src='&<>\"'>&lt;&gt;&amp;&quot;<foo />bar";
    Document doc = Jsoup.parse(h);

    doc.outputSettings().syntax(Syntax.html);
    assertEquals("<!doctype html>\n" +
            "<html>\n" +
            " <head></head>\n" +
            " <body>\n" +
            "  <img async checked src=\"&amp;<>&quot;\">&lt;&gt;&amp;\"\n" +
            "  <foo />bar\n" +
            " </body>\n" +
            "</html>", doc.html());

    doc.outputSettings().syntax(Document.OutputSettings.Syntax.xml);
    assertEquals("<!DOCTYPE html>\n" +
            "<html>\n" +
            " <head></head>\n" +
            " <body>\n" +
            "  <img async=\"\" checked=\"checked\" src=\"&amp;<>&quot;\" />&lt;&gt;&amp;\"\n" +
            "  <foo />bar\n" +
            " </body>\n" +
            "</html>", doc.html());
}

项目：astor 文件：DocumentTest.java

@Test public void testHtmlAndXmlSyntax() {
    String h = "<!DOCTYPE html><body><img async checked='checked' src='&<>\"'>&lt;&gt;&amp;&quot;<foo />bar";
    Document doc = Jsoup.parse(h);

    doc.outputSettings().syntax(Syntax.html);
    assertEquals("<!doctype html>\n" +
            "<html>\n" +
            " <head></head>\n" +
            " <body>\n" +
            "  <img async checked src=\"&amp;<>&quot;\">&lt;&gt;&amp;\"\n" +
            "  <foo />bar\n" +
            " </body>\n" +
            "</html>", doc.html());

    doc.outputSettings().syntax(Document.OutputSettings.Syntax.xml);
    assertEquals("<!DOCTYPE html>\n" +
            "<html>\n" +
            " <head></head>\n" +
            " <body>\n" +
            "  <img async=\"\" checked=\"checked\" src=\"&amp;<>&quot;\" />&lt;&gt;&amp;\"\n" +
            "  <foo />bar\n" +
            " </body>\n" +
            "</html>", doc.html());
}

项目：common 文件：DocumentTest.java

@Test public void testHtmlAppendable() {
    String htmlContent = "<html><head><title>Hello</title></head><body><p>One</p><p>Two</p></body></html>";
    Document document = Jsoup.parse(htmlContent);
    OutputSettings outputSettings = new OutputSettings();

    outputSettings.prettyPrint(false);
    document.outputSettings(outputSettings);
    assertEquals(htmlContent, document.html(new StringWriter()).toString());
}

项目：astor 文件：DocumentTest.java

@Test public void testHtmlAppendable() {
    String htmlContent = "<html><head><title>Hello</title></head><body><p>One</p><p>Two</p></body></html>";
    Document document = Jsoup.parse(htmlContent);
    OutputSettings outputSettings = new OutputSettings();

    outputSettings.prettyPrint(false);
    document.outputSettings(outputSettings);
    assertEquals(htmlContent, document.html(new StringWriter()).toString());
}

项目：astor 文件：DocumentTest.java

@Test public void testHtmlAppendable() {
    String htmlContent = "<html><head><title>Hello</title></head><body><p>One</p><p>Two</p></body></html>";
    Document document = Jsoup.parse(htmlContent);
    OutputSettings outputSettings = new OutputSettings();

    outputSettings.prettyPrint(false);
    document.outputSettings(outputSettings);
    assertEquals(htmlContent, document.html(new StringWriter()).toString());
}

项目：astor 文件：DocumentTest.java

@Test public void testHtmlAppendable() {
    String htmlContent = "<html><head><title>Hello</title></head><body><p>One</p><p>Two</p></body></html>";
    Document document = Jsoup.parse(htmlContent);
    OutputSettings outputSettings = new OutputSettings();

    outputSettings.prettyPrint(false);
    document.outputSettings(outputSettings);
    assertEquals(htmlContent, document.html(new StringWriter()).toString());
}

项目：interview-preparation 文件：App.java

public static void main( String[] args )
   {
    // load html from file
    Document doc = loadHtmlFromFile("index.html", "utf-8"); 
    if(doc == null) {
        LogUtils.d(CLS_NAME, "main", "Document is null");
        return;
    }

    /* BEFORE modification */
    System.out.println("===== BEFORE =====");
    System.out.println(doc.html());

    /* add meta charset utf-8 */
    Element tagMetaCharset = new Element(Tag.valueOf("meta"), "");
    tagMetaCharset.attr("charset", "utf-8");
    // add meta tag to head
    doc.head().appendChild(tagMetaCharset);

    /* create tag <p> for description */
    Element tagPDescription = new Element(Tag.valueOf("p"), "");
    tagPDescription.text("It is a very powerful HTML parser! I love it so much...");
    // add to body
    doc.body().appendChild(tagPDescription);

    /* create tag <p> for author */
    tagPDescription.before("<p>Author: Johnathan Hedley</p>");

    /* add attribute to tag <p> author */
    Element tagPAuthor = doc.body().select("p:contains(Author)").first();
    tagPAuthor.attr("align", "center");

    /* body add class */
    doc.body().addClass("content");
    doc.body().addClass("content");
    doc.body().addClass("content2");

    /* output the final HTML */
    OutputSettings settings = new OutputSettings();
    settings.indentAmount(8);
    doc.outputSettings(settings);
    System.out.println("===== AFTER =====");
    System.out.println(doc.html());
}

项目：concretis 文件：JsoupUserInputSanitizer.java

@Override
public String sanitize(String input, String baseUri, UserInputSanitizerPolicy policy) {
    Whitelist whitelist = getWhitelist(policy);
    OutputSettings outputSettings = new OutputSettings().prettyPrint(false);
    return Jsoup.clean(input, baseUri, whitelist, outputSettings);
}

项目：MathEOS 文件：Patch002.java

@Override
protected boolean apply(DataProfil profil) {
    List<DataTexte> dataTextes = getAllElements(DataTexte.class, profil);
    for(DataTexte dataTexte : dataTextes) {
        org.jsoup.nodes.Document doc = Jsoup.parse(dataTexte.getContenuHTML());
        Elements spanTP = doc.select("span."+JLabelTP.JLABEL_TP);
        for(Element spanElt : spanTP) {
            String spanID = spanElt.attr("id");
            BufferedImage oldImage = dataTexte.putImage(spanID, null);
            int width = oldImage.getWidth(), height = oldImage.getHeight();

            // Get a DOMImplementation.
            DOMImplementation domImpl = GenericDOMImplementation.getDOMImplementation();

            // Create an instance of org.w3c.dom.Document.
            String svgNS = "http://www.w3.org/2000/svg";
            Document document = domImpl.createDocument(svgNS, "svg", null);

            // Create an instance of the SVG Generator.
            SVGGraphics2D svgGenerator = new SVGGraphics2D(document);
            svgGenerator.setSVGCanvasSize(new Dimension(width, height));
            svgGenerator.drawImage(oldImage, 0, 0, null);

            Element svgElement;
            try (StringWriter w = new StringWriter()) {
                svgGenerator.stream(w,true);
                svgElement = Jsoup.parse(w.toString()).outputSettings(new OutputSettings().prettyPrint(false)).select("svg").first();
            } catch (IOException ex) {
                Logger.getLogger(this.getClass().getName()).log(Level.SEVERE, null, ex);
                svgElement = Jsoup.parse("<svg></svg>").select("svg").first();
            }
            Element imgElt = spanElt.select("img").first();
            String id = imgElt.attr("id"), title = imgElt.attr("title");
            svgElement.attr("id", id).attr("width",width+"").attr("height",height+"").attr("title", title);
            String svg = svgElement.outerHtml();
            spanElt.html(svg);
        }
        dataTexte.setContenuHTML(doc.html());
    }
    return true;
}

项目：Server-Application 文件：EmailReceiver.java

private List<Mail> getMails(Message messages[]) throws MessagingException, IOException {
        List<Mail> mails = new ArrayList<Mail>();

        for (int i = 0; i < messages.length; i++) {
            Message msg = messages[i];
            Address[] fromAddress = msg.getFrom();
            Mail mail = new Mail();
            mail.setId(msg.getMessageNumber());
            mail.setFrom(fromAddress[0].toString());
            mail.setSubject(msg.getSubject());
            mail.getToList().addAll(Arrays.asList(parseAddresses(msg.getRecipients(RecipientType.TO))));
            mail.getCcList().addAll(Arrays.asList(parseAddresses(msg.getRecipients(RecipientType.CC))));
            mail.setSendDate(msg.getSentDate().toString());
            mail.setFlags(getFlags(msg));

            String messageContent = "";

            Object msgContent = msg.getContent();
            /* Check if content is pure text/html or in parts */
            if (msgContent instanceof Multipart) {

                Multipart multipart = (Multipart) msgContent;
                for (int j = 0; j < multipart.getCount(); j++) {
                    BodyPart bodyPart = multipart.getBodyPart(j);
                    String disposition = bodyPart.getDisposition();
                    if (disposition != null && (disposition.equals(BodyPart.ATTACHMENT))) {
                        mail.getAttachments().add(getAttachmentDescription(bodyPart));
                    } else if (disposition != null && (disposition.equals(BodyPart.INLINE))) {
                        if (bodyPart.isMimeType("text/*")) {
                            messageContent += (String) bodyPart.getContent();
                        }
                    } else {
                        messageContent += bodyPart.getContent().toString();
                    }
                }
            } else {
                messageContent = messages[i].getContent().toString();
            }

            // escape javascript
//          Document doc = Jsoup.parse(messageContent);
//          doc.removeAttr("script");

             String prettyPrintedBodyFragment = Jsoup.clean(messageContent, "", Whitelist.none().addTags("br", "p"), new OutputSettings().prettyPrint(true));
                // get plain text with preserved line breaks by disabled prettyPrint
            mail.setMessage(Jsoup.clean(prettyPrintedBodyFragment, "", Whitelist.none(), new OutputSettings().prettyPrint(false)));
            mails.add(mail);
        }
        return mails;
    }

项目：cia 文件：DocumentDataPageModContentFactoryImpl.java

@Secured({ "ROLE_ANONYMOUS", "ROLE_USER", "ROLE_ADMIN" })
@Override
public Layout createContent(final String parameters, final MenuBar menuBar, final Panel panel) {
    final VerticalLayout panelContent = createPanelContent();

    final String pageId = getPageId(parameters);

    final DataContainer<DocumentElement, String> documentElementDataContainer = getApplicationManager()
            .getDataContainer(DocumentElement.class);

    final DataContainer<DocumentContentData, String> documentContentDataDataContainer = getApplicationManager()
            .getDataContainer(DocumentContentData.class);

    getApplicationManager()
            .getDataContainer(CommitteeProposalComponentData.class);

    final DocumentElement documentElement = documentElementDataContainer.load(pageId);

    if (documentElement != null) {

        getDocumentMenuItemFactory().createDocumentMenuBar(menuBar, pageId);

        LabelFactory.createHeader2Label(panelContent,DOCUMENT_DATA);


        final List<DocumentContentData> documentContentlist = documentContentDataDataContainer
                .getAllBy(DocumentContentData_.id, pageId);

        if (!documentContentlist.isEmpty()) {

            final Panel formPanel = new Panel();
            formPanel.setSizeFull();

            panelContent.addComponent(formPanel);

            final FormLayout formContent = new FormLayout();
            formPanel.setContent(formContent);

            final String cleanContent = Jsoup.clean(documentContentlist.get(0).getContent(),"", Whitelist.simpleText(), new OutputSettings().indentAmount(4));

            final Label htmlContent = new Label(cleanContent, ContentMode.HTML);

            formContent.addComponent(htmlContent);

            final DocumentWordCountRequest documentWordCountRequest = new DocumentWordCountRequest();
            documentWordCountRequest.setDocumentId(pageId);
            documentWordCountRequest.setMaxResults(MAX_RESULTS);
            documentWordCountRequest.setSessionId(RequestContextHolder.currentRequestAttributes().getSessionId());
            final DocumentWordCountResponse resp = (DocumentWordCountResponse) getApplicationManager().service(documentWordCountRequest);

            if (resp.getWordCountMap() != null) {
                final Label wordCloud = new Label(createWordCloud(resp.getWordCountMap()), ContentMode.HTML);
                formContent.addComponent(wordCloud);
            }

            panelContent.setExpandRatio(formPanel, ContentRatio.GRID);

        }



        panel.setContent(panelContent);
        getPageActionEventHelper().createPageEvent(ViewAction.VISIT_DOCUMENT_VIEW, ApplicationEventGroup.USER, NAME, parameters, pageId);
    }

    return panelContent;

}