Question

如果我尝试读取名为csv_file.csv的CSV文件。问题是，当我用BufferedReader.readLine()读取行时，它会跳过第一行几个月。但是，当我将文件重命名为csv_file.txt时，它会将其读取并且不会跳过第一行。

BufferedReader是否有未记录的“功能”，我不知道？

档案示例：

Months, SEP2010, OCT2010, NOV2010
col1, col2, col3, col4, col5
aaa,,sdf,"12,456",bla bla bla, xsaffadfafda
and so on, and so on, "10,00", xxx, xxx

代码：

FileInputStream stream = new FileInputStream(UploadSupport.TEMPORARY_FILES_PATH+fileName);
BufferedReader br = new BufferedReader(new InputStreamReader(stream, "UTF-8"));
String line = br.readLine();
String months[] = line.split(",");
while ((line=br.readLine())!=null) {
    /*parse other lines*/
}

Answer 1

通常，使用InputStreamReader(InputStream in)构造函数是不好的做法，它使用“默认字符集”。您应该明确指定charset。

但这很难解释你的问题。

Answer 2

我的系统没有区别：

Windows Vista SP2（32位）
NTFS
JDK 1.6.0_17

输出：

Creating C:\workspace\Sandbox\src\data.txt

Reading C:\workspace\Sandbox\src\data.txt
Skipped: 'Months, SEP2010, OCT2010, NOV2010'
First read: 'col1, col2, col3, col4, col5'

Reading C:\workspace\Sandbox\src\data.csv
Skipped: 'Months, SEP2010, OCT2010, NOV2010'
First read: 'col1, col2, col3, col4, col5'

Reading C:\workspace\Sandbox\src\data.txt
Skipped: 'Months, SEP2010, OCT2010, NOV2010'
First read: 'col1, col2, col3, col4, col5'

代码：

import java.io.BufferedReader;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.InputStreamReader;
import java.io.PrintWriter;


public class BuffReadTest {

    public static void main(final String[] args) {
        final String baseFilename = args[0] + "/data";
        try {
            final File txtFile = new File(baseFilename+".txt");
            final File csvFile = new File(baseFilename+".csv");

            if (txtFile.exists())   txtFile.delete();
            if (csvFile.exists())   csvFile.delete();
            createFile(txtFile.getAbsolutePath());

            readFile(txtFile.getAbsolutePath());

            txtFile.renameTo(csvFile);
            readFile(csvFile.getAbsolutePath());

            csvFile.renameTo(txtFile);
            readFile(txtFile.getAbsolutePath());

        } catch (final IOException ex) {
            System.out.println("Exception: "+ex);
            ex.printStackTrace();
        }
    }

    private static void createFile(final String filename)
            throws FileNotFoundException {
        System.out.println("\nCreating "+filename);
        final PrintWriter pw = new PrintWriter(filename);
        pw.println("Months, SEP2010, OCT2010, NOV2010");
        pw.println("col1, col2, col3, col4, col5");
        pw.println("aaa,,sdf,\"12,456\",bla bla bla, xsaffadfafda");
        pw.println("and so on, and so on, \"10,00\", xxx, xxx");
        pw.close();
    }

    private static void readFile(final String filename)
            throws FileNotFoundException, IOException {
        System.out.println("\nReading "+filename);
        final FileInputStream stream = new FileInputStream(filename);
        final BufferedReader br = new BufferedReader(new InputStreamReader(stream));
        final String skipped = br.readLine();
        final String first = br.readLine();

        System.out.println("Skipped: '"+skipped+"'");
        System.out.println("First read: '"+first+"'");
        br.close();
    }

}

Answer 3

嗯，Java会忽略文件的文件扩展名（实际上，通常唯一关心文件扩展名的是Windows）。我猜你会遇到某种新线/非常微妙的编码问题导致你看到的问题。

Answer 4

您的编辑在保存文件时是否做了一些特别的事情？你在Windows上工作吗？（Linux和Windows之间存在一些Newline差异，尽管我从来没有遇到过使用Java的麻烦）

CSV与TXT文件中的Java BufferedReader行为

4 个答案: