Java 类org.jsoup.nodes.DataNode 实例源码

项目：Android-Scrapper 文件：EspnScoreboardParser.java

private void appendGames(Document document) {
    if (document != null) {
        Elements scriptElements = document.getElementsByTag("script");
        Pattern pattern = Pattern.compile("window.espn.scoreboardData[\\s\t]*= (.*);.*window.espn.scoreboardSettings.*");
        for (Element element : scriptElements) {
            for (DataNode node : element.dataNodes()) {
                if (node.getWholeData().startsWith("window.espn.scoreboardData")) {
                    Matcher matcher = pattern.matcher(node.getWholeData());
                    if (matcher.matches()) {
                        EspnJson espnJson = new Gson().fromJson(matcher.group(1), EspnJson.class);
                        teamsList.putAll(espnJson.getTeams());
                    }
                }
            }
        }
    }
}

项目：Android-Scrapper 文件：ExampleUnitTest.java

@Test
public void leagueStatusCheck() throws Exception {
    Document doc = Jsoup.connect("http://www.espn.com/wnba/scoreboard/_/group/50")
            .timeout(60 * 1000)
            .maxBodySize(0)
            .get();
    Elements scriptElements = doc.getElementsByTag("script");
    Pattern pattern = Pattern.compile("window.espn.scoreboardData[\\s\t]*= (.*);.*window.espn.scoreboardSettings.*");
    for (Element element : scriptElements) {
        for (DataNode node : element.dataNodes()) {
            if (node.getWholeData().startsWith("window.espn.scoreboardData")) {
                Matcher matcher = pattern.matcher(node.getWholeData());
                if (matcher.matches()) {
                    Gson gson = new Gson();
                    EspnJson espnJson = new Gson().fromJson(matcher.group(1), EspnJson.class);
                    System.out.println(espnJson.getTeams());
                    assertEquals(false, espnJson.getTeams().isEmpty());
                }
            }
        }
    }
}

项目：Android-Scrapper 文件：ExampleUnitTest.java

@Override
public void run() {
    Document parsedDocument = null;
    try {
        parsedDocument = Jsoup.connect(href).timeout(600 * 1000).get();
    } catch (IOException e) {
        e.printStackTrace();
    }
    Elements scriptElements = parsedDocument.getElementsByTag("script");

    Pattern pattern = Pattern.compile(".*value\":\"(.*)\"\\},\\{\"name.*");
    for (Element element : scriptElements) {
        for (DataNode node : element.dataNodes()) {
            Matcher matcher = pattern.matcher(node.getWholeData().replaceAll("\n", ""));
            if (matcher.matches()) {
                System.out.println(teamCity + "," + StringUtils.capitalize(teamName) + "," + matcher.group(1).toUpperCase());
            }
        }
    }
    Thread.currentThread().interrupt();
}

项目：astor 文件：Cleaner.java

public void head(Node source, int depth) {
    if (source instanceof Element) {
        Element sourceEl = (Element) source;

        if (whitelist.isSafeTag(sourceEl.tagName())) { // safe, clone and copy safe attrs
            ElementMeta meta = createSafeElement(sourceEl);
            Element destChild = meta.el;
            destination.appendChild(destChild);

            numDiscarded += meta.numAttribsDiscarded;
            destination = destChild;
        } else if (source != root) { // not a safe tag, so don't add. don't count root against discarded.
            numDiscarded++;
        }
    } else if (source instanceof TextNode) {
        TextNode sourceText = (TextNode) source;
        TextNode destText = new TextNode(sourceText.getWholeText());
        destination.appendChild(destText);
    } else if (source instanceof DataNode && whitelist.isSafeTag(source.parent().nodeName())) {
      DataNode sourceData = (DataNode) source;
      DataNode destData = new DataNode(sourceData.getWholeData());
      destination.appendChild(destData);
    } else { // else, we don't care about comments, xml proc instructions, etc
        numDiscarded++;
    }
}

项目：astor 文件：Cleaner.java

public void head(Node source, int depth) {
    if (source instanceof Element) {
        Element sourceEl = (Element) source;

        if (whitelist.isSafeTag(sourceEl.tagName())) { // safe, clone and copy safe attrs
            ElementMeta meta = createSafeElement(sourceEl);
            Element destChild = meta.el;
            destination.appendChild(destChild);

            numDiscarded += meta.numAttribsDiscarded;
            destination = destChild;
        } else if (source != root) { // not a safe tag, so don't add. don't count root against discarded.
            numDiscarded++;
        }
    } else if (source instanceof TextNode) {
        TextNode sourceText = (TextNode) source;
        TextNode destText = new TextNode(sourceText.getWholeText());
        destination.appendChild(destText);
    } else if (source instanceof DataNode && whitelist.isSafeTag(source.parent().nodeName())) {
      DataNode sourceData = (DataNode) source;
      DataNode destData = new DataNode(sourceData.getWholeData());
      destination.appendChild(destData);
    } else { // else, we don't care about comments, xml proc instructions, etc
        numDiscarded++;
    }
}

项目：q-mail 文件：HeadCleaner.java

public void head(Node source, int depth) {
    if (skipChildren) {
        return;
    }

    if (source instanceof Element) {
        Element sourceElement = (Element) source;

        if (isSafeTag(sourceElement)) {
            String sourceTag = sourceElement.tagName();
            Attributes destinationAttributes = sourceElement.attributes().clone();
            Element destinationChild = new Element(Tag.valueOf(sourceTag), sourceElement.baseUri(), destinationAttributes);

            destination.appendChild(destinationChild);
            destination = destinationChild;
        } else if (source != root) {
            skipChildren = true;
        }
    } else if (source instanceof TextNode) {
        TextNode sourceText = (TextNode) source;
        TextNode destinationText = new TextNode(sourceText.getWholeText(), source.baseUri());
        destination.appendChild(destinationText);
    } else if (source instanceof DataNode && isSafeTag(source.parent())) {
        DataNode sourceData = (DataNode) source;
        DataNode destinationData = new DataNode(sourceData.getWholeData(), source.baseUri());
        destination.appendChild(destinationData);
    }
}

项目：Android-Scrapper 文件：EspnGameScoreParser.java

private void appendGames(Document scoreBoardDocument, HashMap<Status, List<Competitor>> hashMap) {
    Elements scriptElements = scoreBoardDocument.getElementsByTag("script");
    Pattern pattern = Pattern.compile("window.espn.scoreboardData[\\s\t]*= (.*);.*window.espn.scoreboardSettings.*");
    for (Element element : scriptElements) {
        for (DataNode node : element.dataNodes()) {
            if (node.getWholeData().startsWith("window.espn.scoreboardData")) {
                Matcher matcher = pattern.matcher(node.getWholeData());
                if (matcher.matches()) {
                    EspnJson espnJson = new Gson().fromJson(matcher.group(1), EspnJson.class);
                    hashMap.putAll(espnJson.getStatus());
                }
            }
        }
    }
}

项目：gitplex-mit 文件：JsoupUtils.java

public static void appendReplacement(Matcher matcher, Node node, String replacement) {
    StringBuffer buffer = new StringBuffer();
    matcher.appendReplacement(buffer, "");
    if (buffer.length() != 0)
        node.before(new TextNode(buffer.toString(), node.baseUri()));
    node.before(new DataNode(replacement, node.baseUri()));
}

项目：HfutHelper_Android 文件：LibraryTools.java

public static String getLocationFromText(String text) {
    if (text == null) {
        return null;
    }
    String location = null;
    String strWZxxxxxx = null;
    String strMsg = null;
    Document doc = Jsoup.parse(text);
    Elements scriptTags = doc.getElementsByTag("script");
    String nodeStr = null;
    for (Element tag : scriptTags) {
        for (DataNode node : tag.dataNodes()) {
            nodeStr = node.getWholeData();
            System.out.println("node:" + node);
        }
    }
    for (String str : nodeStr.split("\n")) {
        System.out.println("str:" + str);
        if (str.contains("strWZxxxxxx") && str.contains("|")) {
            strWZxxxxxx = ((str.split("\\|"))[1].split("\""))[0];
            System.out.println(strWZxxxxxx);
        } else if (str.contains("strMsg")) {
            strMsg = (str.split("\""))[1];
            System.out.println(strMsg);
            break;
        }
    }
    if (strMsg != null&&!TextUtils.isEmpty(strMsg)) {
        location = strMsg;
    } else{
        location = strWZxxxxxx;

    }

    return location;
}

项目：ogham 文件：JsoupCssInliner.java

/**
 * Replace link tags with style tags in order to keep the same inclusion
 * order
 *
 * @param doc
 *            the html document
 * @param cssContents
 *            the list of external css files with their content
 */
private void internStyles(Document doc, List<ExternalCss> cssContents) {
    Elements els = doc.select(CSS_LINKS_SELECTOR);
    for (Element e : els) {
        if (!TRUE_VALUE.equals(e.attr(SKIP_INLINE))) {
            String path = e.attr(HREF_ATTR);
            Element style = new Element(Tag.valueOf(STYLE_TAG), "");
            style.appendChild(new DataNode(getCss(cssContents, path), ""));
            e.replaceWith(style);
        }
    }
}

项目：PolymerRenamer 文件：HtmlRenamer.java

@Override
public void head(Node node, int depth) {
  if (node instanceof Element) {
    Element element = (Element) node;
    String tagName = element.tag().getName();
    if (tagName.equals(HtmlTags.POLYMER_ELEMENT.getName())) {
      renameAttributesAttributeValue(element);
    } else if (tagName.equals("script")) {
      insideScriptElement = true;
    } else {
      renameAllAnnotatedEventAttributes(element);
      renameAllAttributeValues(element);
    }
  } else if (node instanceof TextNode) {
    TextNode textNode = (TextNode) node;
    textNode.text(renameStringWithDatabindingDirectives(textNode.getWholeText()));
  } else if (insideScriptElement && node instanceof DataNode) {
    DataNode dataNode = (DataNode) node;
    String js = dataNode.getWholeData();
    try {
      js = JsRenamer.renameProperties(renameMap, js);
    } catch (JavaScriptParsingException e) {
      System.err.println(e);
    }
    dataNode.setWholeData(js);
  }
}

项目：aet 文件：MarkupVisitor.java

@Override
public void visit(Node node) {
  if (node instanceof TextNode || node instanceof Comment || node instanceof DataNode) {
    node.replaceWith(new TextNode(StringUtils.EMPTY, node.baseUri()));
  }
}

项目：touchkit 文件：ApplicationCacheSettings.java

@Override
public void modifyBootstrapPage(BootstrapPageResponse response) {
    Document document = response.getDocument();

    // Add the widgetsetUrl parameter to the bootstrap parameters.
    // This is overridden to avoid adding the naive random query
    // parameter (used by core to avoid caching of js file).

    final VaadinService service = response.getSession().getService();
    final VaadinRequest request = response.getRequest();
    final String staticFilePath = service
            .getStaticFileLocation(request);
    // VAADIN folder location
    final String vaadinDir = staticFilePath + "/VAADIN/";
    // Figure out widgetset
    final UICreateEvent event = new UICreateEvent(request,
            response.getUiClass());
    String widgetset = response.getUIProvider().getWidgetset(event);
    if (widgetset == null) {
        widgetset = request.getService()
                .getConfiguredWidgetset(request);
    }
    // Url for the widgetset
    final String widgetsetUrl = String.format(
            "%swidgetsets/%s/%s.nocache.js", vaadinDir, widgetset,
            widgetset);

    // Update the bootstrap page
    Element scriptTag = document.getElementsByTag("script").last();
    String script = scriptTag.html();
    scriptTag.html("");

    script = script.replace("});", ",\"widgetsetUrl\":\"" + widgetsetUrl
            + "\",\"offlineEnabled\":" + isOfflineModeEnabled() + "});");

    scriptTag.appendChild(new DataNode(script, scriptTag.baseUri()));

    if (isCacheManifestEnabled()) {
        // Add cache manifest attribute to html tag
        document.getElementsByTag("html").attr(
                "manifest",
                vaadinDir + "widgetsets/" + widgetset + "/"
                        + generateManifestFileName(response));
    }
}