SplitByValue Function with IDoc structure

[Former Member]

Hi,

I have an interface where I am trying to create IDocs directly from XI using a Message Mapping (without Java, XSLT, etc.). My input file has a flat structure (including a store field), and the IDoc structure is as follows:

ZSC_DELVRY03 ....................(1)

-

---

IDOC .................................(1)

-

---

EDI_DC40 ......................(1)

-

---

E1EDL20 .......................(1-9999) .......Delivery Header

-

---

ZSC_E1EDL20_ADD....(0-1) ............Extension on header Level

-

---

E1EDL24....................(0-99999)......Delivery Item

I need to create a new E1EDL20 (Header) segment for each store (on change of store in the input file, which is sorted). I have been able to split the file by store using the Standard Function SplitByValue on the store field, however there is a problem.

In the XI mapping documentation it says that a maximum of min occurs target fields(structures) can be created using SplitByValue. Therefore, since the E1EDL20 Segment has min occurs = max occurs = 1, only the first E1EDL20 segment for the first store is created.

Has anyone found a way round this? Is there any way of changing the XSD definition for the Idoc in XI?

Thanks,

--Lee Tomlinson

[Former Member]

Hello Lee,

according to your description of the problem, your E1EDL20-Segment might occur 1-9999 times? So minOccurs=1 and maxOccurs=9999?

In this case you do not need the SplitByValue function. Simply map a source segment or a source node that represents your store (probably the store field) to E1EDL20. It should be created as often as there are store fields in your source message.

Best regards

Joachim

[Former Member]

I do not think I can have explained my problem properly. Here is what I am actually trying to do:

The structure of my input file is as follows where the store is the first 4 chars and item code is the following 6 chars:

Store

| Item code

0123000001...(other fields)

0123000002...(other fields)

0123000003...(other fields)

0456000004...(other fields)

0456000005...(other fields)

0456000006...(other fields)

0456000007...(other fields)

0456000008...(other fields)

0789000009...(other fields)

0789000010...(other fields)

I do not want to create an E1EDL20 segment for each of the records in the input file, only when the store changes. I want to do this, so that the items for a particular store all appear in the lower level segments of the appropriate E1EDL20 segment.

IDOC

-

-

---

E1EDL20-WERKS....... = 0123

-

-

---

E1EDL24-MATNR.... = 000001

-

-

---

E1EDL24-MATNR.... = 000002

-

-

---

E1EDL24-MATNR.... = 000003

-

-

---

E1EDL20-WERKS........= 0456

-

-

---

E1EDL24-MATNR.... = 000004

-

-

---

E1EDL24-MATNR.... = 000005

-

-

---

E1EDL24-MATNR.... = 000006

-

-

---

E1EDL24-MATNR.... = 000007

-

-

---

E1EDL24-MATNR.... = 000008

.

.

. etc.

The problem is that because the E1EDL20 segment has MIN occurs 1, when you use splitByValue, the field (or in this case structure - E1EDL20) will only be created a MAXIMUM of Min Occurs times (i.e.- once).

I am looking for a way of getting round this issue, as I cannot change the structure of the IDoc.

Thanks a lot,

--Lee Tomlinson