Iterate problem, 1 type: group, 2 type: segment

Johnny,

I am just curious if you have run the inbound message(s) through the HRL Testing tool and if so were the repetitions where you expected and did the data look correct?

Johnny,

I really don’t have the time to delve deeply into this today and I will be in Houston for the next week. I will try to look more deeply into it as time permits.

In the meantime I think you have the 02_17 defined as repeating in two places. The primary HRL for the actual sub-HRL and again in the sub-HRL.

I would verify that is correct. If it is not (I suspect it is not) then that may be confusing the parser thus the ITERATE not functioning as expected.

I suspect you only want the repetition specified in the parent HRL at the point it is referencing the sub-HRL.

Another suggestion is to echo out the %g and %s counters to validate the values are what you expect at the appropriate time.

Johnny,

Use a $ in front of counters – like $%g1 when referencing as a Source or Destination (and thus affecting Tcl).

That should allow you to echo out the value.

Hi Johnny,

When I look at the testing results of your HRL, none of the testmessages are showing that the HRL ‘Types_02_17’ is being filled. So imho the HRL definition isn’t working, although it looks valid to me.

Do I understand correctly that your ‘Types_02_17’ consists of one Type_02, followed by zero or more Type_17? Maybe you could try to SET Type_02 inside the HRL Types_02_17 to No Repeat.

But you have to get the HRL working correctly first.

I’m not sure if you get this HRL inside an HRL to work; I haven’t seen it mentioned on this forum before and I don’t have a working example of it either.

I have defined HRL inside HRL so basically that does work.

I do agree that the place to concentrate is on the message construct definition versus the actual data.

Again, when I get back to my office I should have more time to delve into this.

In the meantime, something I do is to make a very skeleton Xlate just enough to validate the structure before aI flesh out the rest. That allows me to validate the structure before I have too much ‘noise’ to contend with.

Ok, based on these latest testingtool results you sent, I think I got it.

I am going to talk about you first xlate (xlate.ppt), where you have an iterate within an iterate. So not that later xlate_45.ppt, where you  have only one iteration.

In the first iterate, you use %g1. The iterate within, you use %s1. This is ok. However, for the outbound PID segments, you are also using %s1 and this is where things go wrong. Why? Because either %g1 or %s1 can increase, meaning a new PID segment should be output. You can clearly see this in the last testingtool result of the HRL.

So you will need to define and maintain your own counter (%g9; I am using a group counter because you are repeating a group) to keep track of the number of PID segments that are output.

Before the iteration, set %g9 to 0:

Then in all the PID segments, use %g9 as your variable, instead of %s1 you are using now.

At the end of writing a complete PID-segment, you need to increase %g9:

That should do the trick. Hope this works.

I’m going to try to explain.

If you look at the test_tool_results that you sent yesterday, you will see in message 3 the following lines:

So in this message the HRL Types_02_17 is iterating, with only 1 Type_02 and optionally a Type_17 following.

Message 5 shows this:

Here you have only one Type_02_17 with many Type_02 iterating within it.

Therefore you can also have a combination of this: several iterations of Type_02_17 and within each iteration an iteration of many Type_02.

I think I can conclude though that each occurrence of a Type_02 (optionally followed by a Type_07) should result in a PID-segment, correct?

In that case, in your last xlate you need to place the iteration of Type_17 AND incrementing %g9 inside the iteration of Type_02. Practically: select all lines of your xlate from the position in slide 5 up until the position in slide 6 and then move this selection ‘up’ in the translation 2 steps. These lines should now be inside the iteration of Type_02, so after each Type_02 a PID segment is created.