we can use run compononent in the parameters tab for commandline give unix AWK command for skipping the specified record num

awk '{if(NR!=1  && NR!=2) print $0}' this we can use for first record

or else we can use onemore component read seperated value in the parameters tab u can specify skipheader as true

This is upto my knowledge

Thanks

Bhanu

If the input dml has record type, simply do FBE will do.

If the input dml has no record type, but with "n" as End of Record, do a wc -l up front and pass the result back to the graph will do the trick.

Otherwise, you need to assign a sequence number (next_in_sequence) using Reformat, also capture the log port for the record count.  Then:

Header: sequence number=1

Do a join with the source data (say in0) and from the record count capture from the previous Reformat log port (say, in1).

Trailer: in0.sequencenumber=in1.reccount

guys,

 to read header ,detail,trailer from input file separately ,do validatins for these,

use conditional dml,where we can separate detail from header and trailer

for validations u have to use reformat with count :3(out0:header,Out1:detail,Out2:trailer.)

if any doubts further let me know

cheers

mukund