Pentaho CSV 文件输入步骤 - 附件中的逗号分隔符

Question

我对 Pentaho 6.1 CSV 文件输入步骤有奇怪的问题。

我在 CSV 文件中有以下值：

"2016-09-16","test","","test","test, SL","[test]","test",39.000000,0.000000,0.000000,0.000000,0.000000,0.00000000000.00000000000. ,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.0000000,0.0000000,0,0000000,0.0000000

我在 CSV 文件输入步骤中使用：

Delimiter: ,
Enclosure: "

预览选项完美运行，但是当我尝试将该内容导入数据库表并执行转换时，出现以下错误：

2016/09/23 22:04:46 - 写入 log.0 - 2016/09/23 22:04:46 - 写入 log.0 - 2016/09/23 22:04:46 - 写入 log.0 - 在 org.pentaho.di.core.row.value.ValueMetaBase.convertStringToNumber(ValueMetaBase.java:803) 2016/09/23 22:04:46 - 写入 log.0 - 在 org.pentaho.di.core。 row.value.ValueMetaBase.getNumber(ValueMetaBase.java:1631) 2016/09/23 22:04:46 - 写入 log.0 - ... 7 more 2016/09/23 22:04:46 - 写入 log .0 - 由：org.pentaho.di.core.exception.KettleValueException：2016/09/23 22:04:46 - 写入 log.0 - 请求字符串：无法将字符串转换为数字：非数字字符位置1处找到的值[SL”， “[测试]”， “测试”，38.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000， 0.000000,0.000000,0.000000,0.000000,0.000000

我认为它有以下值“test, SL”（第 5 列）的问题。 我不明白为什么，因为预览选项有效。

Answer 1

我试过你的输入集，不管你信不信，我对嵌入的逗号没有问题。 问题似乎是数字字段的转换。 我将它们设置为Number数据类型并摆弄了掩码，但唯一使它起作用的是指定修剪类型“左”。 当我这样做时，一切都很好。