OpenCsv 用 BeanToCsv + HeaderColumnNameTranslateMappingStrategy 写错列名

Question

我正在使用 opencsv 3.6 从 java bean 创建一个 csv 文件。

首先，我尝试了这段代码：

import com.opencsv.CSVReader;
import com.opencsv.CSVWriter;
import com.opencsv.bean.BeanToCsv;
import com.opencsv.bean.HeaderColumnNameTranslateMappingStrategy;

import java.io.FileReader;
import java.io.FileWriter;
import java.io.IOException;

import java.util.ArrayList;
import java.util.HashMap;
import java.util.Map;

public class CustomBean {

    private String name;
    private String surname;

    public CustomBean(String n, String s) {
        this.name = n;
        this.surname = s;
    }

    public void setName(String name) {
        this.name = name;
    }

    public String getName() {
        return name;
    }

    public void setSurname(String surname) {
        this.surname = surname;
    }

    public String getSurname() {
        return surname;
    }

    public static void main(String[] args) {
        Map<String,String> mapping = new HashMap<String,String>();
        mapping.put("COLUMN1","name");
        mapping.put("COLUMN2","surname");

        HeaderColumnNameTranslateMappingStrategy<CustomBean> strategy = new HeaderColumnNameTranslateMappingStrategy<CustomBean>();
        strategy.setType(CustomBean.class);
        strategy.setColumnMapping(mapping);

        ArrayList<CustomBean> customUsers = new ArrayList<CustomBean>();
        customUsers.add(new CustomBean("Kobe","Bryant"));

        BeanToCsv<CustomBean> bean = new BeanToCsv<CustomBean>();

        try {            
            CSVWriter writer = new CSVWriter(new FileWriter("testOut.csv"));
            bean.write(strategy, writer, customUsers);
            writer.close();
        } catch (IOException e) {
            e.printStackTrace();
        }

    }
}

但是我遇到了以下错误：

Exception in thread "main" java.lang.RuntimeException: Error writing CSV !
    at com.opencsv.bean.BeanToCsv.write(BeanToCsv.java:74)
    at test.CustomBean.main(CustomBean.java:63)
Caused by: java.lang.NullPointerException
    at com.opencsv.bean.HeaderColumnNameTranslateMappingStrategy.getColumnName(HeaderColumnNameTranslateMappingStrategy.java:45)
    at com.opencsv.bean.HeaderColumnNameMappingStrategy.findDescriptor(HeaderColumnNameMappingStrategy.java:112)
    at com.opencsv.bean.BeanToCsv.processHeader(BeanToCsv.java:103)
    at com.opencsv.bean.BeanToCsv.write(BeanToCsv.java:69)
    ... 1 more

发生这种情况是因为在opencsv源代码中getColumnName方法中HeaderColumnNameTranslateMappingStrategyclass有下面一行：

return col < header.length ? columnMapping.get(header[col].toUpperCase()) : null;

因此，header 为空。这是真的，事实上这个 class 是 HeaderColumnNameMappingStrategy class 的子 class 包含 header 变量（String[] 类型）从未初始化。

我在这个 class 中找到的唯一有用的方法是 captureHeader，但不幸的是它需要一个 CSVReader 作为输入。

出于这个原因，我创建了一个空的 csv 文件：

COLUMN1,COLUMN2

并且我在 try/catch 块的开头添加了以下行：

CSVReader reader = new CSVReader(new FileReader("testIn.csv"));
strategy.captureHeader(reader);

这种方式（我真的不喜欢，因为我必须创建一个csv文件）我也不例外，但是在结果csv文件中列的名称不遵循映射策略：

"name","surname"
"Kobe","Bryant"

两个问题：

如何获得预期的结果，即 csv 文件中的正确列名？
有办法不使用 CSVReader class?

Answer 1

查看 BeanToCsv 的源代码，processHeader(...) 方法对提供的 headers 没有任何作用。您唯一的选择是创建自定义策略（以避免 CSVReader ）和自定义 BeanToCsv 如下

 

public class CustomBean {
   ...
   public static void main(String[] args) {
     ...
     HeaderColumnNameTranslateMappingStrategy strategy = new <b>CustomStrategy</b><CustomBean>();
     strategy.setType(CustomBean.class);
     strategy.setColumnMapping(mapping);
     ...

     BeanToCsv bean = new <b>CustomBeanToCsv</b><CustomBean>();
     ...
   }

   static class CustomStrategy<T> extends HeaderColumnNameTranslateMappingStrategy {

    @Override
    public void setColumnMapping(Map columnMapping) {
        super.setColumnMapping(columnMapping);
        header = new String[columnMapping.size()];
        int i = 0;
        for (Map.Entry entry : columnMapping.entrySet()) {
            header[i] = entry.getKey().toUpperCase();
            i++;
        }
    }

     public String[] getHeader() {
         return header;
     }
   }

   static class CustomBeanToCsv<T> extends BeanToCsv {
     @Override
     protected String[] processHeader(MappingStrategy mapper) throws IntrospectionException {
         if (mapper instanceof CustomStrategy) {
             return ((CustomStrategy) mapper).getHeader();
         } else {
             return super.processHeader(mapper);
         }
     }
   }
}

Answer 2

我已经使用 openCSV 五年了，而且我仍在学习有关它的知识。问题是 HeaderColumnNameMappingStrategy 和 HeaderColumnNameTranslateMappingStrategy 是为 CsvToBean 制作的。它需要一个文件来读取文件头以查看 reader 应该从哪里读取。

对于 BeanToCsv class 使用 ColumnPositionMappingStrategy class。你给它你正在映射的 class 和你想要映射的列的列表，它会为你做剩下的。

这是我写的一个有效的小测试方法。

public void createUsingBeanToCsv(int numRecords, FileWriter fos) {
    List<SmallRecord> smallRecords = new ArrayList<>(numRecords);
    for (int i = 0; i < numRecords; i++) {
        smallRecords.add(SmallRecordGenerator.createRecord(i));
    }

    BeanToCsv<SmallRecord> beanToCsv = new BeanToCsv<>();
    ColumnPositionMappingStrategy<SmallRecord> strategy = new ColumnPositionMappingStrategy<>();
    strategy.setType(SmallRecord.class);
    String[] columns = new String[]{"bigDecimal_1", "name_1", "intNumber_1"};
    strategy.setColumnMapping(columns);

    beanToCsv.write(strategy, fos, smallRecords);

}

OpenCsv 用 BeanToCsv + HeaderColumnNameTranslateMappingStrategy 写错列名

OpenCsv writes wrong column names with BeanToCsv + HeaderColumnNameTranslateMappingStrategy

java

csv

opencsv