articledomparser.java

用java实现的一个bbs的portal

/**
 * Created by IntelliJ IDEA.
 * User: Administrator
 * Date: Jul 17, 2003
 * Time: 11:14:38 AM
 * To change this template use Options | File Templates.
 */
package Classification;

import org.w3c.dom.Document;
import org.w3c.dom.NodeList;
import org.w3c.dom.Element;
import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.parsers.DocumentBuilder;
import java.util.Vector;

import Kernel.Article;

public class ArticleDomParser {

    private Document doc;
    private String FilePath;
    private String FileName;
    private Vector ArticleVec = new Vector();
    private boolean isContinue = true;

    public ArticleDomParser(String filename){
        this.FileName = filename;
        this.FilePath = ArticleBatch.NewArticlePath + filename;
        System.out.println(FilePath);
        if (freshFromFile())
            doParse();
        else isContinue = false;
    }

    public boolean isContinue() {
        return isContinue;
    }

    private boolean freshFromFile(){
        try  {
            DocumentBuilderFactory  docbuilderfactory  =  DocumentBuilderFactory.newInstance();
            DocumentBuilder  docbuilder  =  docbuilderfactory.newDocumentBuilder();
            doc  =  docbuilder.parse(FilePath);
            doc.normalize();
        }  catch  (Exception  ex)  {
            return false;
        }
        return true;
    }

    private void doParse() {
        try  {
            Element BBSInfo = (Element)doc.getElementsByTagName("BBSInfo").item(0);
            String BBS = BBSInfo.getElementsByTagName("BBS").item(0).getFirstChild().getNodeValue();
            String Board = BBSInfo.getElementsByTagName("Board").item(0).getFirstChild().getNodeValue();
            Element Article = (Element)BBSInfo.getElementsByTagName("Article").item(0);
            String Author = Article.getElementsByTagName("Author").item(0).getFirstChild().getNodeValue();
            String Alias = " ";
            if (Article.getElementsByTagName("Alias").item(0).getFirstChild() != null)
                Alias = Article.getElementsByTagName("Alias").item(0).getFirstChild().getNodeValue();
            String Title = " ";
            if (Article.getElementsByTagName("Title").item(0).getFirstChild() != null)
                Title = Article.getElementsByTagName("Title").item(0).getFirstChild().getNodeValue();
            String Content = " ";
            if (Article.getElementsByTagName("Content").item(0).getFirstChild() != null)
                Content = Article.getElementsByTagName("Content").item(0).getFirstChild().getNodeValue();
            String Date = Article.getElementsByTagName("Date").item(0).getFirstChild().getNodeValue();

            String IP = "";
            if (Article.getElementsByTagName("IP").item(0).getFirstChild() != null)
                IP = Article.getElementsByTagName("IP").item(0).getFirstChild().getNodeValue();

            Alias = tidyString(Alias);
            Title = tidyString(Title);
            Content = tidyString(Content);
            ArticleVec.add(new Article(true,BBS,Board,Author,Alias,Title,Content,Date,IP,FileName));
            NodeList REList = doc.getElementsByTagName("RE");
            for (int i=0;i<REList.getLength();i++){
                Element RENode = (Element)REList.item(i);
                String RE_Author = RENode.getElementsByTagName("Author").item(0).getFirstChild().getNodeValue();
                String RE_Alias = " ";
                if (RENode.getElementsByTagName("Alias").item(0).getFirstChild() != null)
                    RE_Alias = RENode.getElementsByTagName("Alias").item(0).getFirstChild().getNodeValue();
                String RE_Content = " ";
                if (RENode.getElementsByTagName("Content").item(0).getFirstChild() != null)
                    RE_Content = RENode.getElementsByTagName("Content").item(0).getFirstChild().getNodeValue();
                String RE_Date = RENode.getElementsByTagName("Date").item(0).getFirstChild().getNodeValue();
                String RE_IP = RENode.getElementsByTagName("IP").item(0).getFirstChild().getNodeValue();
                ArticleVec.add(new Article(false,BBS,Board,RE_Author,RE_Alias,Title,RE_Content,RE_Date,RE_IP,FileName));
            }
        }  catch  (Exception  ex)  {
            ex.printStackTrace();
        }
    }

    public Vector getArticleVec(){
        return this.ArticleVec;
    }

    private String tidyString(String str){
        str = str.replaceAll("#","＃");
        str = str.replaceAll("@","◎");
        str = str.replaceAll("'","‘");
        str = str.replaceAll("%","％");
        return str;
    }

    public static void main(String args[]){
        ArticleDomParser domparser = new ArticleDomParser("BDWM_SecondHand_2003-07-17 19-25-18.xml");
        System.out.println(domparser.ArticleVec.size());
    }
}