As explained in the following chapter, Type Mappings and Data Conversion, the use of nested and repeating data types can result in output files with large numbers of columns and rows. The -minLevel and -maxLevel command-line options are used to create vertical slices from an input data file.

The following example specification demonstrates how these options work to reduce the output:

   A ::= SEQUENCE {
      a INTEGER,
      b SEQUENCE OF SEQUENCE {
         bb VisibleString,
         cc CHOICE {
            aaa INTEGER,
            bbb SEQUENCE OF BOOLEAN
         }
      }
   }      
      

Without using any command-line filtering options, the output columns will look like this:

   a,bb,aaa,bbb      
      

The innermost SEQUENCE OF type will cause a full tuple to be added to the CSV file for each message. If the bbb element were repeated ten times, the outer elements would be duplicated ten times for each BOOLEAN.

If, in the same message, the outer SEQUENCE OF (that is, the b element) were repeated three times, the outer INTEGER, a, would be repeated 30 times. This kind of duplication may be unnecessary depending on the content of interest, so the minLevel and maxLevel options may be used to control the output.

The duplication of data at the outer level may be controlled using the minLevel option. If for example, the minimum level were set to one (-minLevel 1), the outer INTEGER would be eliminated:

   bb,aaa,bbb      
      

The duplication of data in the inner levels may be controlled using the maxLevel option. If, for example, the maxLevel were set to one (-maxLevel 1), the inner CHOICE would be eliminated:

   a,bb      
      

By combining the options, we can reduce the output to a single column of data (-minLevel 1 -maxLevel 1):

   bb      
      

In this way the data of interest may be isolated in the input messages and the output considerably reduced.