Problem to transform Flat file to Data Type Structurated and map to IDOC

Question

Hi all,
I have a file to idoc scenario.
The information is like this:
1#!445#!AI12#!1#!20070214#!DVXXXXR#!201#!31GINHG876#!#!
2#!#!ETC
3#!000000000030008888#!#!3000#!#!10#!#!20070215
4#!dades45#!b#!c#!d#!e#!f#!g
5#!pos5
where 1, 2, ... = key segments for the file adapter and #! are the field separator.
this must go to a Data Type with substructures like this:
(xml)
.1
......2
.3
......4
......5
where 2 is inside 1 and 4-5 inside 3.
The reason of this is we can get unbounded replays of 3/4/5 for one header.
But XI reads the information like if all segments are headers.
.1
.2
.3
.4
.5
So the IDOC is bad created.
Where is the issue?  At Message Mapping all substructures are mapped with his default and at TEST option it works fine.
Thanks in advance for your help.
best regards
Message was edited by:
        Federico Martin
Message was edited by:
        Federico Martin

Answer 1

Dear people,
according to license problems at client, it's impossible to get the convert agent.
so now we come back to try to solve with normal mapping.
the next step is try to convert it using 2 message mappings: dt (flat xml) to dt (structurated) and this one to the idoc.
i suppose it can be done adding program at interface mapping.
I created the no hiereachical Data Type for load the information and it's fine. But i am unable to get the first mapping OK, because XI has problems with unbounded segments.
If you agreed and has time, i copy here the information and his structure.
I receive (#! defines separation and key segment fields are 1,2,3,4,5):
1#!445#!AH02#!1#!20070214#!DVPOSTER#!201#!31GINHGIN0#!#!
2#!#!ETC
3#!000000000030008888#!#!3000#!#!10#!#!20070215
4#!dades45#!b#!c#!d#!e#!f#!g
3#!000000000030008888#!#!3000#!#!10#!#!20070215
3#!000000000030008888#!#!3000#!#!10#!#!20070215
4#!dades45#!b#!c#!d#!e#!f#!g
5#!pos5
5#!pos5
5#!pos5
The destination must be:
CHASE A
1 (1..1)
....2 (1..1)
3 (1..unbounded)
....4 (0..1)
3 (1..unbounded)
3 (1..unbounded)
....4 (0..1)
5 (1..unbounded)
5 (1..unbounded)
5 (1..unbounded)
CHASE B
Or, in another case with 5 as child of 3 (like example in last mail)
1#!445#!AH02#!1#!20070214#!DVPOSTER#!201#!31GINHGIN0#!#!
2#!#!ETC
3#!000000000030008888#!#!3000#!#!10#!#!20070215
4#!dades45#!b#!c#!d#!e#!f#!g
5#!pos5
3#!000000000030008888#!#!3000#!#!10#!#!20070215
5#!pos5
3#!000000000030008888#!#!3000#!#!10#!#!20070215
4#!dades45#!b#!c#!d#!e#!f#!g
5#!pos5
1 (1..1)
....2 (1..1)
3 (1..unbounded)
....4 (0..1)
....5 (1..unbounded)
3 (1..unbounded)
....5 (1..unbounded)
3 (1..unbounded)
....4 (0..1)
....5 (1..unbounded)
3 (1..unbounded)
....4 (0..1)
....5 (1..unbounded)
Questions:
How i do for map Data Type (no hierachical) -to-> Data Type (hierachical) and  Data Type (hierachical) -to-> IDOC? Is any loop or context object required?
Thank you in advance for your help and sorry the long thread.