Hello Kolusu Due to the holiday, I have only studied the sample JCL you provided in the past few days.
The JCL is of great help to me. Thanks for your help! I failed to describe our needs clearly last time. We hope to use dfsort to compare today’s operlog with yesterday’s. There are two kinds of messages in operlog. 1. single-line message 2. multi-line message The problem is that the number of lines in a multi-line message is variable. To simplify the problem, we make two operlog. operlog is FBM and LRECL=133 How is a duplicate identified? The 'key' is the position from 58 to 133. The duplicate is identified by the 'key' Below is operlog(1) . There are 6 messages in the operlog operlog(1): 1 20 58 133 ----------------------------+------------------------------------------------------+-------------------------.................-------+ M 4040000 CUK1 18179 16:02:44.45 STC00280 00000090 HZS0002E A1 E B1 M 4040000 CUK1 18179 16:03:44.45 STC00280 00000090 HZS0002E A1 E B1 MR0000000 CUK1 18179 16:08:04.53 AWRK08 00000090 $HASP636 A2A2 DR B2B2 DR C2C2 ER D2D2 MR0000000 CUK1 18179 16:08:05.53 AWRK08 00000090 $HASP636 A2A2 DR B2B2 DR C2C2 ER D2D2 N 4008000 CUK1 18179 16:08:06.12 STC00389 00000090 AOF869I A3A3A3 N 8000000 CUK1 18179 16:33:58.48 00000090 DSI064A A4A4A4A4 S B4B4B4B4 --------------------------------------------------------------------------------------------------------------------------------------------- Question1 : Because the number of lines in a multi-line message is variable, Could you help us to get the following result (operlog(2)) by selecting the first duplicate of multi-line messages or single-line messages ? operlog(2): 1 20 58 133 ----------------------------+------------------------------------------------------+-------------------------.................-------+ M 4040000 CUK1 18179 16:02:44.45 STC00280 00000090 HZS0002E A1 E B1 MR0000000 CUK1 18179 16:08:04.53 AWRK08 00000090 $HASP636 A2A2 DR B2B2 DR C2C2 ER D2D2 N 4008000 CUK1 18179 16:08:06.12 STC00389 00000090 AOF869I A3A3A3 N 8000000 CUK1 18179 16:33:58.48 00000090 DSI064A A4A4A4A4 S B4B4B4B4 S C4C4C4C4 --------------------------------------------------------------------------------------------------------------------------------------------- Now there are 4 messages in operlog(2) Now we made the other operlog(3). please see it below . There are two messages in the operlog(3) operlog(3): 1 20 58 133 ----------------------------+------------------------------------------------------+-------------------------.................-------+ M 4040000 CUK1 18179 16:02:44.45 STC00280 00000090 HZS0002E A1 E B1 MR0000000 CUK1 18179 16:08:04.53 AWRK08 00000090 $HASP636 A2A2 DR B2B2 DR C2C2 ER D2D2 --------------------------------------------------------------------------------------------------------------------------------------------- Question 2 : Because the number of lines in a multi-line message is variable,could you help us to compare the operlog(2) with the operlog(3) to get the operlog(4) The compare's 'key' is also the position from 58 to 133. operlog(4): 1 20 58 133 ----------------------------+------------------------------------------------------+-------------------------.................-------+ N 4008000 CUK1 18179 16:08:06.12 STC00389 00000090 AOF869I A3A3A3 N 8000000 CUK1 18179 16:33:58.48 00000090 DSI064A A4A4A4A4 S B4B4B4B4 S C4C4C4C4 --------------------------------------------------------------------------------------------------------------------------------------------- Question3: If the number of lines in a multi-line message is fixed, the manual tells us how to merge multiple lines into a single line? Because the number of lines in a multi-line message is actually variable, Could you help us to use ICETOOL operator RESIZE to merge multiple lines into a single line for operlog(2) and operlog(3). After merging multiple lines into a single line for operlog(2) and operlog(3), we could compare them. Question4 : After comparing them , how to split a single line into multiple lines? Thanks a lot! Jason Cai ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN