[ https://issues.apache.org/jira/browse/FLINK-17818?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17192945#comment-17192945 ]
Danish Amjad edited comment on FLINK-17818 at 9/28/20, 1:14 PM: ---------------------------------------------------------------- [~twalthr] Sorry for the late reply. I have created a pull request. https://github.com/apache/flink/pull/13367 cc: [~aljoscha] was (Author: damjad): [~twalthr] Sorry for the late reply. I have created a pull request. https://github.com/apache/flink/pull/13367 > CSV Reader with Pojo Type and no field names fails > -------------------------------------------------- > > Key: FLINK-17818 > URL: https://issues.apache.org/jira/browse/FLINK-17818 > Project: Flink > Issue Type: Bug > Components: API / DataSet > Affects Versions: 1.10.1 > Reporter: Danish Amjad > Priority: Minor > Labels: pull-request-available > Original Estimate: 0.5h > Remaining Estimate: 0.5h > > When a file is read with a CSVReader and a POJO is specified and the filed > names are not specified, the output is obviously not correct. The API is not > throwing any error despite a null check inside the API. i.e. > {code} > Preconditions.checkNotNull(pojoFields, "POJO fields must be specified (not > null) if output type is a POJO."); > {code} > The *root cause* of the problem is that the _fieldNames_ argument is a > variable argument and the variable is 'empty but not null' when not given any > value. So, the check passes without failing. > I am not sure if it is a feature of the API or an actual bug because there is > a test checking the NPE when a null is passed. -- This message was sent by Atlassian Jira (v8.3.4#803005)