I am currently working on TOS_BD_6.1.1.
I am using a tSqoopImport component with java API and it works fine, creating the file I need on the cluster.
However, when I try to custom my file, using "Advanced Settings -> Additionnal arguments", I am not able to produce all of the following requirements :
- fields separated with semicolon
- fields enclosed into double quote
- escape character \ to be used to escape double quote
The additionnal arguments I defined are :
"codegen.output.delimiters.field" -> ";"
"codegen.output.delimiters.enclose" -> "\""
"codegen.output.delimiters.escape" -> "\\"
As an example, if I have a row of 3 columns with the following data : banana | ban"ana" | ban"ana
I want the file created to be like that : "banana";"ban\"ana\"";"ban\"ana"
But only the semicolon works and I got : banana;ban"ana";ban"ana
You can see my tSqoopImport configuration on the 2 screenshots attached.
Am I using them wrongly? Or is it a known bug?
I saw this similar topic https://community.talend.com/t5/Design-and-Development/tsqoopimport-ignores-additional-arguments-D-m... but the OP didn't answer.
Thanks for your time.
The escape character \ to escape double quote is actually working with the described configuration, but I am still not able to enclose my fields, no matter what character I choose (double quote with "\"", or another one, like "£".
By default, what character are escaped ? double quote, single quote, etc..., ?
I have to escape them because I may encountered double quote, and special character like \t , \n in the data I extract.