电子报纸采集程序

使用JSoup+CSSPath从网上采集电子报纸,完整代码见search-demo,下面是主要的接口,抽象类和实现。通过对不同电子报纸的网站结构进行分析,可以抽象出一些通用的模式,统一的接口,通用的抽象类,可扩展的实现。

 

1、接口

/**
 *报纸采集器
 * @author 杨尚川
 */
public interface PaperCollector {
    /**
     * 下载当日报纸,一个文件对应一个版面
     * @return 报纸
     */
    List<File> collect();
    /**
     * 下载指定日期的报纸,一个文件对应一个版面
     * @param date 指定日期
     * @return 报纸
     */
    List<File> collect(Date date);
}

 

2、抽象类

 

/**
 *报纸采集器抽象类,通用采集功能实现
 * @author 杨尚川
 */
public abstract class AbstractPaperCollector implements PaperCollector{
    protected final Logger LOG = LoggerFactory.getLogger(getClass());

    @Override
    public List<File> collect() {
        return collect(new Date());
    }
    /**
     * 根据下载链接提取文件夹名称
     * @param href 下载链接
     * @return 文件夹名称
     */
    protected abstract String getPath(String href);
    /**
     * 根据下载链接提取文件名称
     * @param href 下载链接
     * @return 文件名称
     */
    protected abstract String getFile(String href);
    protected List<File> downloadPaper(List<String> hrefs){
        final List<File> files = new ArrayList<>();
        List<Thread> ts = new ArrayList<>();
        LOG.info("报纸有"+hrefs.size()+"个版面需要下载:");
        for(final String href : hrefs){   
            Thread t = new Thread(new Runnable(){
                @Override
                public void run() {
                    File file = downloadPaper(href);
                    if(file != null){
                        files.add(file);
                    }
                }                
            });
            t.start();
            ts.add(t);
        }
        for(Thread t : ts){
            try {
                t.join();
            } catch (InterruptedException ex) {
                LOG.error("下载报纸出错:",ex);
            }
        }
        return files;
    }
    protected File downloadPaper(String href){
        try{
            LOG.info("下载报纸:"+href);
            String path = getPath(href);
            LOG.debug("报纸保存目录:"+path);
            String file = getFile(href);
            LOG.debug("报纸保存文件:"+file);
            File dir = new File(path);
            if(!dir.exists()){
                LOG.debug("创建目录:"+dir.getAbsolutePath());
                dir.mkdirs();
            }
            File absoluteFile = new File(path, file);
            LOG.debug("报纸保存绝对路径:"+absoluteFile.getAbsolutePath());
            Tools.copyFile(new URL(href).openStream(), absoluteFile);
            LOG.info("报纸下载成功:"+href);
            LOG.info("报纸成功保存到:"+absoluteFile.getAbsolutePath());
            return absoluteFile;
        }catch(IOException e){
            LOG.error("报纸下载失败:"+e);
        }
        return null;
    }    
    protected void run() {
        //今天
        List<File> files = collect();
        int i = 1;
        for(File file : files){
            LOG.info((i++)+" : " + file.getAbsolutePath());
        }
        //昨天
        Date date = new Date();
        date.setTime(System.currentTimeMillis()-24*3600*1000);
        files = collect(date);
        i = 1;
        for(File file : files){
            LOG.info((i++)+" : " + file.getAbsolutePath());
        }
        //前天
        date = new Date();
        date.setTime(System.currentTimeMillis()-2*24*3600*1000);
        files = collect(date);
        i = 1;
        for(File file : files){
            LOG.info((i++)+" : " + file.getAbsolutePath());
        }
    }
}

 

3、采集新华日报

 

/**
 * 新华日报
 * @author 杨尚川
 */
public class XHRBPaperCollector extends AbstractPaperCollector{
    private static final String paperName = "新华日报";
    private static final String paperPath = "http://xh.xhby.net/newxh/";
    private static final String url = paperPath+"html/";
    private static final String hrefPrefix = paperPath+"page/1/";
    private static final String start = "node_2.htm";
    private static final String pdfCssQuery = "html body table tbody tr td table tbody tr td table tbody tr td table tbody tr td div table tbody tr td a";
    private static final SimpleDateFormat sf = new SimpleDateFormat("yyyy-MM/dd/"); 
    
    @Override
    public List<File> collect(Date date) {
        List<String> hrefs = new ArrayList<>();
        try {
            LOG.debug("url: "+url);
            String paper = url + sf.format(date) + start;
            LOG.debug("paper: "+paper);
            Document document = Jsoup.connect(paper).get();
            
            LOG.debug("pdfCssQuery: " + pdfCssQuery);
            Elements elements = document.select(pdfCssQuery);
            for(Element element : elements){
                String href = element.attr("href");
                if(href != null && href.endsWith(".pdf")){
                    LOG.debug("报纸链接:"+href);
                    href = href.replace("../../../", "");
                    LOG.debug("报纸链接:"+href);
                    hrefs.add(paperPath+href);
                }else{
                    LOG.debug("不是报纸链接:"+href);
                }
            }            
        } catch (IOException ex) {
            LOG.error("采集出错",ex);
        }
        return downloadPaper(hrefs);
    }
    @Override
    protected String getPath(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        attrs = attrs[0].split("-");
        StringBuilder str = new StringBuilder();
        str.append(paperName)
            .append(File.separator)
            .append(attrs[0])
            .append("-")
            .append(attrs[1])
            .append(File.separator)
            .append(attrs[2]);
        return str.toString();
    }
    @Override
    protected String getFile(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        String file = attrs[1]+".pdf";
        return file;
    }
    public static void main(String[] args) {
        new XHRBPaperCollector().run();
    }
}

 

 4、采集楚天都市报

 

/**
 * 楚天都市报
 * @author 杨尚川
 */
public class CTDSBPaperCollector extends AbstractPaperCollector{
    private static final String paperName = "楚天都市报";
    private static final String host = "http://ctdsb.cnhubei.com/";
    private static final String paperPath = host+"ctdsb/";
    private static final String url = host+"html/ctdsb/";
    private static final String hrefPrefix = paperPath;
    private static final String start = "index.html";
    private static final String pdfCssQuery = "html body center table tbody tr td table tbody tr td table tbody tr td table tbody tr td div table tbody tr td.info3 a";
    private static final SimpleDateFormat sf = new SimpleDateFormat("yyyyMMdd/"); 
    
    @Override
    public List<File> collect(Date date) {
        List<String> hrefs = new ArrayList<>();
        try {
            LOG.debug("url: "+url);
            String paper = url + sf.format(date) + start;
            LOG.debug("paper: "+paper);
            Document document = Jsoup.connect(paper).get();
            
            LOG.debug("pdfCssQuery: " + pdfCssQuery);
            Elements elements = document.select(pdfCssQuery);
            int count=0;
            for(Element element : elements){
                String text = element.text();
                if(text != null && text.startsWith("第")){
                    LOG.debug("报纸文本:"+text);
                    count++;
                }else{
                    LOG.debug("不是报纸文本:"+text);
                }
            }
            //有的版面缺失,而文件名是顺序递增的
            for(int i=1; i<=count; i++){
                String seq = Integer.toString(i);
                if(i<10){
                    seq="0"+seq;
                }
                hrefs.add(paperPath + sf.format(date) + "page_"+seq+".jpg");
            }
        } catch (IOException ex) {
            LOG.error("采集出错",ex);
        }
        return downloadPaper(hrefs);
    }
    @Override
    protected String getPath(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        StringBuilder str = new StringBuilder();
        str.append(paperName)
            .append(File.separator)
            .append(attrs[0].substring(0, 4))
            .append("-")
            .append(attrs[0].substring(4, 6))
            .append(File.separator)
            .append(attrs[0].substring(6, 8));
        return str.toString();
    }
    @Override
    protected String getFile(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        String file = attrs[1].split("_")[1];
        return file;
    }
    public static void main(String[] args) {
        new CTDSBPaperCollector().run();
    }
}

 

5、采集京九晚报

/**
 * 京九晚报
 * @author 杨尚川
 */
public class JJWBPaperCollector extends AbstractPaperCollector{
    private static final String paperName = "京九晚报";
    private static final String paperPath = "http://epaper.cnsq.com.cn/jjwb/";
    private static final String url = paperPath+"html/";
    private static final String hrefPrefix = paperPath+"page/10/";
    private static final String start = "node_11.htm";
    private static final String pdfCssQuery = "html body table tbody tr td table tbody tr td table tbody tr td table tbody tr td div table tbody tr td a";
    private static final SimpleDateFormat sf = new SimpleDateFormat("yyyy-MM/dd/"); 
    
    @Override
    public List<File> collect(Date date) {
        List<String> hrefs = new ArrayList<>();
        try {
            LOG.debug("url: "+url);
            String paper = url + sf.format(date) + start;
            LOG.debug("paper: "+paper);
            Document document = Jsoup.connect(paper).get();
            
            LOG.debug("pdfCssQuery: " + pdfCssQuery);
            Elements elements = document.select(pdfCssQuery);
            for(Element element : elements){
                String href = element.attr("href");
                if(href != null && href.endsWith(".pdf")){
                    LOG.debug("报纸链接:"+href);
                    href = href.replace("../../../", "");
                    LOG.debug("报纸链接:"+href);
                    hrefs.add(paperPath+href);
                }else{
                    LOG.debug("不是报纸链接:"+href);
                }
            }            
        } catch (IOException ex) {
            LOG.error("采集出错",ex);
        }
        return downloadPaper(hrefs);
    }
    @Override
    protected String getPath(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        StringBuilder str = new StringBuilder();
        str.append(paperName)
            .append(File.separator)
            .append(attrs[0])
            .append(File.separator)
            .append(attrs[1]);
        return str.toString();
    }
    @Override
    protected String getFile(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        String file = attrs[2]+".pdf";
        return file;
    }
    public static void main(String[] args) {
        new JJWBPaperCollector().run();
    }
}

 

 

 6、采集信息时报

/**
 * 信息时报
 * @author 杨尚川
 */
public class XXSBPaperCollector extends AbstractPaperCollector{
    private static final String paperName = "信息时报";
    private static final String host = "http://informationtimes.dayoo.com/";
    private static final String paperPath = host+"page/1019/";
    private static final String url = host+"html/";
    private static final String hrefPrefix = paperPath;
    private static final String start = "node_1019.htm";
    private static final String pdfCssQuery = "html body#content div.container div.leftcolumn div.leftcolumncontent div.pagebuttontwo div.con p.right span.dfive a";
    private static final String subCssQuery = "html body#listcontent div.container div.rightcolumn div.subcbga div.listcontent div#all_article_list.list h4 span.left a";
    private static final String contentCssQuery = "html body div.container div.leftcolumn div.tbga div.bbga div.cbga div.left div.pagepicture div map area";
    private static final SimpleDateFormat sf = new SimpleDateFormat("yyyy-MM/dd/"); 
    
    @Override
    public List<File> collect(Date date) {
        List<String> hrefs = new ArrayList<>();
        try {
            LOG.debug("url: "+url);
            String paper = url + sf.format(date) + start;
            LOG.debug("paper: "+paper);
            Document document = Jsoup.connect(paper).get();
            
            //1、找到子报纸
            LOG.debug("subCssQuery: " + subCssQuery);
            Elements elements = document.select(subCssQuery);
            for(Element element : elements){
                String text = element.text();
                String href = element.attr("href");
                if(text != null && text.contains(":") && href != null && href.endsWith(".htm")){
                    String subPaperURL = url + sf.format(date) + href;
                    LOG.debug("子报纸文本:"+text+" , "+href);
                    LOG.debug("subPaperURL:"+subPaperURL);
                    //2、找到内容页面
                    LOG.debug("contentCssQuery: " + contentCssQuery);
                    Elements contentElements = Jsoup.connect(subPaperURL).get().select(contentCssQuery);
                    for(Element contentElement : contentElements){
                        String h = contentElement.attr("href");
                        if(h != null && h.startsWith("content_") && h.endsWith(".htm")){
                            String contentURL = url + sf.format(date) + h;
                            LOG.debug("contentURL:"+contentURL);
                            //3、找PDF
                            LOG.debug("pdfCssQuery: " + pdfCssQuery);
                            Elements pdfElements = Jsoup.connect(contentURL).get().select(pdfCssQuery);
                            for(Element pdfElement : pdfElements){
                                String pdf = pdfElement.attr("href");
                                if(pdf != null && pdf.endsWith(".pdf")){
                                    LOG.debug("报纸链接:"+pdf);
                                    pdf = pdf.replace("../../../", "");
                                    LOG.debug("报纸链接:"+pdf);
                                    hrefs.add(host+pdf);
                                }else{
                                    LOG.debug("不是报纸链接:"+pdf);
                                }
                            }
                            //有多个content,选择一个即可
                            break;
                        }
                    }
                }else{
                    LOG.debug("不是子报纸文本:"+text+" , "+href);
                }
            }
        } catch (IOException ex) {
            LOG.error("采集出错",ex);
        }
        return downloadPaper(hrefs);
    }
    @Override
    protected String getPath(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        StringBuilder str = new StringBuilder();
        str.append(paperName)
            .append(File.separator)
            .append(attrs[0])
            .append(File.separator)
            .append(attrs[1]);
        return str.toString();
    }
    @Override
    protected String getFile(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        String file = attrs[2]+".pdf";
        return file;
    }
    public static void main(String[] args) {
        new XXSBPaperCollector().run();
    }
}

 

 

7、采集羊城晚报

/**
 * 羊城晚报
 * @author 杨尚川
 */
public class YCWBPaperCollector extends AbstractPaperCollector{
    private static final String paperName = "羊城晚报";
    private static final String paperPath = "http://www.ycwb.com/ePaper/ycwb/";
    private static final String url = paperPath+"html/";
    private static final String hrefPrefix = paperPath+"images/";
    private static final String start = "node_2081.htm";
    private static final String pdfCssQuery = "html body div.cbody div.areaL div.box div.conBox2 div div.xx h2 em a.px12";
    private static final SimpleDateFormat sf = new SimpleDateFormat("yyyy-MM/dd/"); 
    
    @Override
    public List<File> collect(Date date) {
        List<String> hrefs = new ArrayList<>();
        try {
            LOG.debug("url: "+url);
            String paper = url + sf.format(date) + start;
            LOG.debug("paper: "+paper);
            Document document = Jsoup.connect(paper).get();
            
            LOG.debug("pdfCssQuery: " + pdfCssQuery);
            Elements elements = document.select(pdfCssQuery);
            for(Element element : elements){
                String href = element.attr("href");
                if(href != null && href.endsWith(".pdf")){
                    LOG.debug("报纸链接:"+href);
                    href = href.replace("../../../", "");
                    LOG.debug("报纸链接:"+href);
                    hrefs.add(paperPath+href);
                }else{
                    LOG.debug("不是报纸链接:"+href);
                }
            }            
        } catch (IOException ex) {
            LOG.error("采集出错",ex);
        }
        return downloadPaper(hrefs);
    }
    @Override
    protected String getPath(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        StringBuilder str = new StringBuilder();
        str.append(paperName)
            .append(File.separator)
            .append(attrs[0])
            .append(File.separator)
            .append(attrs[1]);
        return str.toString();
    }
    @Override
    protected String getFile(String href) {
        String path = href.replace(hrefPrefix, "");
        String[] attrs = path.split("/");
        String file = attrs[2]+".pdf";
        return file;
    }
    public static void main(String[] args) {
        new YCWBPaperCollector().run();
    }
}

 

 项目托管在Github

 

电子报纸采集程序

上一篇:mysql加入?管理员


下一篇:PL/SQL简介