【JAVA】获取网页的单词数并输出到CSV

从网页获取单词，并统计数量，然后输出到CSV文档中。

代码是第一次写，逻辑很混乱，而且功能也不是很完善...

尝试了几个网页，仅对单纯的文档类网页有效.....

虽然是个练习，不过练习的挺失败的...传上来权当纪念。

下面是部分代码。

package com.test.dataCount;

import java.io.BufferedReader;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.net.URL;
import java.net.URLConnection;
import java.util.ArrayList;
import java.util.List;
import java.util.regex.Pattern;

import com.test.bean.DataBean;
import com.test.bean.ParmBean;

public class DataCount {

	private static Pattern SPLIT_PATTERN =
		Pattern.compile("<(\".*?\"|'.*?'|[^'\"])*?>");

	public DataBean execute(ParmBean parmBean) throws Exception {

		DataBean dataBean = new DataBean();

		URL url = null;
		URLConnection URLconn = null;

		InputStream inputStream = null;
		InputStreamReader inputSR = null;
		BufferedReader bufferedReader = null;

		try{

			url = new URL(parmBean.getUrl());
			URLconn = url.openConnection();

			inputStream = URLconn.getInputStream();
			inputSR = new InputStreamReader(inputStream, "UTF-8");
			bufferedReader = new BufferedReader(inputSR);

			List<String> dataCountsList = this.getCounts(bufferedReader, parmBean.getTarget());
			dataBean.setDataList(dataCountsList);

		} finally {

			if (bufferedReader != null) {
				bufferedReader.close();
			}
			if (inputSR != null) {
				inputSR.close();
			}
			if (inputStream != null) {
				inputStream.close();
			}
		}

		return dataBean;

	}

	private List<String> getCounts(BufferedReader bufferedReader, String target) throws Exception {

		List<String> list = new ArrayList<String>();
		String line = null;

		while ((line = bufferedReader.readLine()) != null) {
			if (line.trim().isEmpty()) {
				continue;
			}
			String[] strSplit = SPLIT_PATTERN.split(line);
			for (int i = 0; i < strSplit.length; i++) {
				if (strSplit[i].startsWith(target)) {
					list.add(strSplit[i]);
				} else {
					continue;
				}
			}
		}
		return list;
	}

}

上面这个class用来获取并统计词数。

package com.test.fileWrite;

import java.io.File;
import java.io.FileOutputStream;
import java.io.OutputStreamWriter;
import java.io.PrintWriter;
import java.util.List;

import com.test.bean.DataBean;

public class CSVFileWriter {

	private File file = null;
	protected FileOutputStream fileOutputStream = null;
	private OutputStreamWriter outputStreamWriter = null;
    protected PrintWriter printWriter = null;


    public void execute(DataBean dataBean) throws Exception {

    	init(dataBean);

    	execProcess(dataBean);

    	close();

    }

    public void init(DataBean dataBean) throws Exception {

        this.file = new File(dataBean.getFileName());
		System.out.println("ファイルのパス： " + this.file.getAbsolutePath());

        this.fileOutputStream = new FileOutputStream(this.file);
		this.outputStreamWriter = new OutputStreamWriter(this.fileOutputStream, "UTF-8");
        this.printWriter = new PrintWriter(this.outputStreamWriter);
    }

    public void execProcess(DataBean dataBean) {

    	List<String> list = dataBean.getDataList();

		printWriter.println("No.\tItem\t");

		for (int i = 0; i < list.size(); i++) {

			printWriter.println((i + 1) + "\t" + list.get(i));

		}
    }

    public void close() throws Exception {

		if (this.outputStreamWriter != null) {

			this.outputStreamWriter.flush();
			this.outputStreamWriter.close();
		}

		if (this.fileOutputStream != null) {

			this.fileOutputStream.flush();
			this.fileOutputStream.close();
		}

		if (this.printWriter != null) {

			this.printWriter.flush();
			this.printWriter.close();
		}
    }

}

上面这段代码用来输出到CSV。

main函数和两个Bean省略。前者调用这两个class，后者声明单词，数字以及URL等成员变量。

全程没有注释自己回过头来看也是费劲的很.（主要是代码很乱..）等有时间再整理一下。