Java 将PubMed GEO DataSets中dataset类型的检索结果转化为Excel

检索结果举例
为进行生信分析，需要对检索结果转化为Excel并且以标签作为列。首先将检索结果Send to File转化成TXT格式，如下：


通过下面的程序进一步转为Excel(使用jxl包)：

源代码：

/*
 * To change this license header, choose License Headers in Project Properties.
 * To change this template file, choose Tools | Templates
 * and open the template in the editor.
 */
package ncbi;

import java.io.BufferedReader;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.FileReader;
import java.io.IOException;
import jxl.Workbook;
import jxl.write.Label;
import jxl.write.WritableSheet;
import jxl.write.WritableWorkbook;

/**
 *
 * @author wgyklh
 */
public class NCBI {

    /**
     * @param args the command line arguments
     * @throws java.lang.Exception
     */
    public static void main(String[] args) throws Exception {
        // TODO code application logic here
        File in = new File("C:\\Users\\Lenovo\\Desktop\\gds_result.txt");
        File out = new File("C:\\Users\\Lenovo\\Desktop\\gds.result.xls");
        WritableWorkbook wwb = Workbook.createWorkbook(out);
        WritableSheet dataset = wwb.createSheet("DataSet", 0);
        WritableSheet series = wwb.createSheet("Series", 1);
        WritableSheet platform= wwb.createSheet("Platform", 2);
        WritableSheet sample= wwb.createSheet("Sample", 3);

        dataset.addCell(new Label(0, 0, "Number"));
        dataset.addCell(new Label(1, 0, "Title"));
        dataset.addCell(new Label(2, 0, "Introduction"));
        dataset.addCell(new Label(3, 0, "Organism"));
        dataset.addCell(new Label(4, 0, "Type"));
        dataset.addCell(new Label(5, 0, "Platform"));
        dataset.addCell(new Label(6, 0, "Series"));
        dataset.addCell(new Label(7, 0, "Sample"));
        dataset.addCell(new Label(8, 0, "FTP download"));
        dataset.addCell(new Label(9, 0, "Accession"));
        dataset.addCell(new Label(10, 0, "ID"));

        series.addCell(new Label(0, 0, "Number"));
        series.addCell(new Label(1, 0, "Title"));
        series.addCell(new Label(2, 0, "Introduction"));
        series.addCell(new Label(3, 0, "Organism"));
        series.addCell(new Label(4, 0, "Type"));
        series.addCell(new Label(5, 0, "Platform"));
        series.addCell(new Label(6, 0, "Sample"));
        series.addCell(new Label(7, 0, "FTP download"));
        series.addCell(new Label(8, 0, "Accession"));
        series.addCell(new Label(9, 0, "ID"));
        
        
        BufferedReader reader = null;
        String tempString = null;
        int number = 1;
        int index1 = 0;
        int index2 = 0;
        int index3 = 0;
        try {
            reader = new BufferedReader(new FileReader(in));

            for (int n = 1; n <= 32; n++) {
                tempString = reader.readLine();
                tempString = reader.readLine();
                System.out.println(tempString);
                dataset.addCell(new Label(0, number, String.valueOf(number)));
                dataset.addCell(new Label(1, number, tempString.substring(3)));
                tempString = reader.readLine();
                dataset.addCell(new Label(2, number, tempString));
                tempString = reader.readLine();
                dataset.addCell(new Label(3, number, tempString.substring(10)));
                tempString = reader.readLine();
                dataset.addCell(new Label(4, number, tempString.substring(6)));
                tempString = reader.readLine();
                index1 = tempString.indexOf("Series");
                index2 = tempString.indexOf("Sample");
                System.out.println(index1 + "   " + index2);
                dataset.addCell(new Label(5, number, tempString.substring(10, index1 - 1)));
                dataset.addCell(new Label(6, number, tempString.substring(index1 + 8, index1 + 16)));
                dataset.addCell(new Label(7, number, tempString.substring(index1 + 16, index2 - 1).trim()));
                tempString = reader.readLine();
                dataset.addCell(new Label(8, number, tempString.substring(14)));
                tempString = reader.readLine();
                index3 = tempString.indexOf("ID");
                dataset.addCell(new Label(9, number, tempString.substring(20, 27)));
                dataset.addCell(new Label(10, number, tempString.substring(index3 + 4)));
                number++;
            }
            reader.close();
            wwb.write();
            wwb.close();
        } catch (FileNotFoundException e) {
        } catch (IOException e) {
        } finally {
            if (reader != null) {
                try {
                    reader.close();
                } catch (IOException e) {
                }
            }
        }

    }
}