The output includes an empty element for any headers not found in the parsed file. To skip some fields, simply
omit
them from the list of Custom Object Fields in the Event Source properties.
The output still includes file information.
The process object name is derived as <sourceName>Content, to represent a collection of record objects and other details.
When you have several delimited content or fixed width event sources, each of them uses its own custom content and record objects. The record object uses the format <sourceName>Record.
Field names are converted to NCName format, to remove any prohibited characters from the header names and ensure they are valid process object field names.

Age,Born,User Name
22,1987,Bob Smith
45,1999,Bill White

<MyUserFileContent>
     <fileInfo>
          <!-- for FTP/SFTP only -->
          <host>127.0.0.1</host>        
        
          <lastModified> 2015-04-29T14:37:35.448Z </lastModified>
          <dir>DataFiles</dir>
          <name> users </name>
          <path>/DataFiles/users.csv </path>
          <fullName> users.csv </fullName>
          <ext> csv </ext>
          <size> 78 </size>
     </fileInfo>
     <record>
          <Age> 22 </Age>
          <Born> 1987 </Born>
          <Salary/>
          <User_Name> Bob Smith </User_Name>
     </record>
     <record>
          <Age> 45 </Age>
          <Born> 1999 </Born>
          <Salary/>
          <User_Name> Bill White </User_Name>
     </record>
     <totalRowsCount>3</totalRowsCount>
</MyUserFileContent>