I have a csv file with the following content:
And connect to it with:
When I read the metadata of the result, the type of the first column is LONGVARCHAR(-1) in stead of INTEGER(4)
This regression happened between version 4.2.194 and 4.2.197
Because according to one customer's suggestion, the default value of maxScanRows has been setted to -1. You should use jdbc:csv:/.?_CSV_Separator=\x3b;_CSV_Header=true;delayedClose=0;maxScanRows=0.