使用Spring Batch导入CSV到MySQL遇FlatFileParseException错误

错误原因 #

触发IncorrectTokenCountException的核心原因是DelimitedLineTokenizer的配置冲突：

• 通过setNames定义了24个字段名，Spring Batch据此预期每行解析出24个token
• 同时用setIncludedFields指定只提取7列，实际解析后仅得到7个token
• 预期数量与实际数量不匹配，导致报错

解决方案 #

方案一：匹配字段名数量与提取列数

调整setNames的字段名数组，仅保留需要映射的7个字段，与setIncludedFields指定的列数完全一致：

private LineMapper<Player> getLineMapper() {
    
    DefaultLineMapper<Player> lineMapper = new DefaultLineMapper<>();
    
    DelimitedLineTokenizer lineTokenizer = new DelimitedLineTokenizer();
    
    // 字段名数量与includedFields的列数一一对应
    lineTokenizer.setNames(new String[] {"playerID", "birthYear", "birthCity", "nameFirst", "nameLast", "retroID", "bbrefID"});
    lineTokenizer.setIncludedFields(new int[] {0, 1, 6, 13, 14, 22, 23});
    
    BeanWrapperFieldSetMapper<Player> fieldSetMapper = new BeanWrapperFieldSetMapper<>();
    fieldSetMapper.setTargetType(Player.class);
    
    lineMapper.setLineTokenizer(lineTokenizer);
    lineMapper.setFieldSetMapper(fieldSetMapper);
    
    return lineMapper;
}

方案二：利用空字段名自动跳过列

如果希望保留原setNames的写法（用空字符串标记跳过的列），需移除setIncludedFields配置，DelimitedLineTokenizer会自动忽略字段名为空的列：

private LineMapper<Player> getLineMapper() {
    
    DefaultLineMapper<Player> lineMapper = new DefaultLineMapper<>();
    
    DelimitedLineTokenizer lineTokenizer = new DelimitedLineTokenizer();
    
    // 空字符串对应的列会被自动跳过
    lineTokenizer.setNames(new String[] {"playerID","birthYear","","","","","birthCity","","","","","","","nameFirst","nameLast","","","","","","","","retroID","bbrefID"});
    // 移除setIncludedFields配置
    
    BeanWrapperFieldSetMapper<Player> fieldSetMapper = new BeanWrapperFieldSetMapper<>();
    fieldSetMapper.setTargetType(Player.class);
    
    lineMapper.setLineTokenizer(lineTokenizer);
    lineMapper.setFieldSetMapper(fieldSetMapper);
    
    return lineMapper;
}

额外验证 #

从错误日志的input内容来看，People2.csv的第二行（已跳过第一行表头）确实包含24个逗号分隔的列，符合配置要求，无需调整文件内容。

内容的提问来源于stack exchange，提问作者Aryan Singh