Search in sources :

Example 1 with OrFilter

use of org.htmlparser.filters.OrFilter in project laogewen by wen4034.

the class HtmlParserTool method extracLinks.

public static Set<String> extracLinks(String url, LinkFilter filter, String... validate) {
    Set<String> links = Sets.newHashSet();
    try {
        URL realurl = new URL(url);
        URLConnection connection = realurl.openConnection();
        connection.setRequestProperty("User-Agent", "Mozilla/4.0 (compatible; MSIE 5.0; Windows NT; DigExt)");
        connection.setReadTimeout(100000);
        connection.setConnectTimeout(100000);
        Parser parser = new Parser(connection);
        parser.setEncoding("UTF-8");
        // 过滤<frame>标签的filter,用来提取frame标签的src属性
        NodeFilter frameFilter = new NodeFilter() {

            @Override
            public boolean accept(Node node) {
                if (node.getText().startsWith("frame src=")) {
                    return true;
                } else {
                    return false;
                }
            }
        };
        OrFilter linkFilter = new OrFilter(new NodeClassFilter(LinkTag.class), frameFilter);
        NodeList list = parser.extractAllNodesThatMatch(linkFilter);
        for (int i = 0; i < list.size(); i++) {
            Node tag = list.elementAt(i);
            if (tag instanceof LinkTag) {
                LinkTag linkTag = (LinkTag) tag;
                String linkurl = linkTag.getLink();
                if (filter.accept(linkurl, validate)) {
                    links.add(linkurl);
                }
            } else {
                String fram = tag.getText();
                int start = fram.indexOf("src=");
                fram = fram.substring(start);
                int end = fram.indexOf(" ");
                if (end == -1) {
                    end = fram.indexOf(">");
                }
                String frameUrl = fram.substring(5, end - 1);
                if (filter.accept(frameUrl, validate)) {
                    links.add(frameUrl);
                }
            }
        }
    } catch (Exception e) {
        System.out.println(url + "链接失败");
        e.printStackTrace();
    }
    return links;
}
Also used : NodeClassFilter(org.htmlparser.filters.NodeClassFilter) LinkTag(org.htmlparser.tags.LinkTag) Node(org.htmlparser.Node) NodeList(org.htmlparser.util.NodeList) OrFilter(org.htmlparser.filters.OrFilter) URL(java.net.URL) URLConnection(java.net.URLConnection) Parser(org.htmlparser.Parser) NodeFilter(org.htmlparser.NodeFilter)

Aggregations

URL (java.net.URL)1 URLConnection (java.net.URLConnection)1 Node (org.htmlparser.Node)1 NodeFilter (org.htmlparser.NodeFilter)1 Parser (org.htmlparser.Parser)1 NodeClassFilter (org.htmlparser.filters.NodeClassFilter)1 OrFilter (org.htmlparser.filters.OrFilter)1 LinkTag (org.htmlparser.tags.LinkTag)1 NodeList (org.htmlparser.util.NodeList)1