I use spring-batch 2.1.8 release. Depending if the batch is executed under windows or unix, there’s a different EOL character. On unix, the file generated only has CR – on winsows, there are both CRLF.
I configured my writter chunk so that it encodes the output resource has a windows one, like so :
<property name="encoding" value="cp1252"/>
But no sir. Unix still generates with only CR. It’s in the context of a j2ee classic app with spring, controllers and models. So I can write code after the batch is done, before prompting the file to download for the user.
What am I doing wrong?
You could try setting the lineSeparator property of the writer (assuming you are using a FlatFileItemWriter). The javadocs indicate it is using the System property line.separator for the default.