message mapping for E1EDPT1/E1EDPT2/TDLINE split and assign in proper context of E1EDPT2

[Former Member]

hi all,

i've a source structure and target structure which looks like this below

        <L_PO1>

          <S_PO1>

            <D_330>30</D_330>

            <D_355>BD</D_355>

            <D_235>BP</D_235>

            <D_234>MAT1</D_234>

          </S_PO1>

          <L_PID>

            <S_PID>

              <D_349>F</D_349>

              <D_352>description which can have length 140 characters max</D_352>

            </S_PID>

          </L_PID>

          <L_PID>

            <S_PID>

              <D_349>F</D_349>

              <D_352>another description for the same MAT1 which also can have max 140 char</D_352>

            </S_PID>

          </L_PID>

        </L_PO1>

The target structure is a IDOC orders and it should look like this below after the mapping of item level segment

<E1EDP01 SEGMENT="1">

<MENGE>30</MENGE>

<MENEE>BD</MENEE>

<E1EDP19 SEGMENT="1">

<QUALF>001</QUALF>

<IDTNR>MAT1</IDTNR>

</E1EDP19>

<E1EDPT1 SEGMENT="1">

<TDID>Z001</TDID>

<TSSPRAS>E</TSSPRAS>

<E1EDPT2 SEGMENT="1">

<TDLINE>take first 70 characters of the description</TDLINE> "if the description is greater than 70 characters

<TDFORMAT>/</TDFORMAT>

</E1EDPT2>

<E1EDPT2 SEGMENT="1">

<TDLINE>remaining 70 characters or less from the description</TDLINE> "create a new TDLINE with a new context of E1EDPT2.

<TDFORMAT>/</TDFORMAT>

</E1EDPT2>

<E1EDPT2 SEGMENT="1">

<TDLINE>this description is not greater than 70 characters</TDLINE> " no need to split because it's not more than 70 characters.

<TDFORMAT>/</TDFORMAT>

</E1EDPT2>

</E1EDPT1>

</E1EDP01>

so, basically, I have to split description into 70 characters if its more than 70 characters and assign to a new TDLINE field.

howz it doable?

thx

mike

[Harish · Accepted solution]

Hi Michael,

you need to use UDF for this req. Please refer the UDF code given by Amit in below thread

regards,

Harish

[Former Member · Accepted solution]

Harish,

I am not really looking for splitting. Splitting can be done using standard function Node pool fragment single value but once split I am not sure now to assign the values to E1EDPT2 and TDLINE to repeat for each and every value and thats where I need help to be precise Should have made my ask clear before.

thx

mike

[engswee · Accepted solution]

Edit: Swapped E1EDPT1 and E1EDPT2 mapping

Hi Michael

Try the following UDF of type All Values of Queue. It has two output fields.

public void genText(String[] a, ResultList parent, ResultList line, Container container) throws StreamTransformationException {
  int limit = 70;
  for(int i = 0; i < a.length; i++) {
   if(a[i].equals(ResultList.CC)) {
    line.addContextChange();
    parent.addContextChange();
   } else {
    String text = a[i];
    while(!text.isEmpty()) {
     parent.addValue("");
     if(text.length() > limit) {     
      line.addValue(text.substring(0, limit));
      line.addContextChange();
      text = text.substring(limit);
     } else {
      line.addValue(text);
      break;
     }
    }
   }
  }
} 

Following is the mapping logic for the E1EDPT1 E1EDPT2 and TDLINE fields.

Sample output (I've not included the other ORDERS05 fields)

Rgds

Eng Swee

Message was edited by: Eng Swee Yeoh

[Former Member · Accepted solution]

Hi Eng,

thanks a lot for the reply. But this is not exactly I want. I want all the description for the same item under one E1EDP01 node. With your UDF, it is adding context for each and every E1EDPT2 value which is not what I want. For the above example, I want all the description to come under same E1EDP01 node under multiple E1EDPT2 /TDLINE under a single E1EDPT1 node. Makes sense?

thx

mike

[former_member184720 · Accepted solution]

I think you just need to add remove context after the UDF when you map it E1EDPT2  segment.

Your mapping should be like this-

D_352 -> UDF -> Remove context ->E1EDPT2

D_352 -> UDF ->E1EDPT2/TDLINE

[Former Member · Accepted solution]

hi Harish,

The udf is written to add parent context (E1EDPT2) and if we remove the context for E1EDPT2, then it won't work properly because E1EDPT2 if it is assigned 2 null values for example before CC, it will repeat only 2 times instead of 3 times if the description is 2 and one of them is more than 70 characters. makes sense? And if we pass all the descriptin to E1EDPT2 without any context change, it will repeat more than 3 times when it was supposed to repeat only 3 times. This will create erroneous result.

thx

mike

[Former Member · Accepted solution]

Eng should be able to help is my hope. Any one else before Eng replies.

thx

mike

[engswee · Accepted solution]

Mike

I was lazy and didn't get around to load a full 850 as the source structure, so couldn't really see the minor context difference.

The change is real easy, just comment out a line.

I didn't map the rest of the fields, but you should be able to take care of the rest.

Output

Rgds

Eng Swee

[engswee · Accepted solution]

Actually wait. That does not work if there are more than one PO1. Give me a few minutes. I'll sort it out... Hahaha....

[former_member184720 · Accepted solution]

Ooops.. Deleted my reply.. just saw your reply and i see that both does same thing

Don't know logic for other header node's creation to give a try..

Message was edited by: Hareesh Gampa

[Former Member · Accepted solution]

yes hahaha

thx

mike

[engswee · Accepted solution]

Ok this should be it, hopefully (crossing fingers!)

Change the execution type to "All Values of a Context"

  int limit = 70;  
  for(int i = 0; i < a.length; i++) {  
    String text = a[i];  
    while(!text.isEmpty()) {  
     parent.addValue("");  
     if(text.length() > limit) {       
      line.addValue(text.substring(0, limit));  
      text = text.substring(limit);  
     } else {  
      line.addValue(text);  
      break;  
     }  
    }   
  }

Slight changes to the mapping flow

- Change context of D_352 to L_PO1

- Add SplitByValue to second output of UDF before TDLINE

I've tested with two occurrences of L_PO1 and it should work fine.

[Former Member · Accepted solution]

Very neat Eng. I will test and mark it as correct answer soon. Appreciate your effort.

thx

mike

[Former Member · Accepted solution]

absolutely awesome. thanks a lot Eng. Wouldn't have done without your help.

thx

mike

[engswee · Accepted solution]

Brilliant! Really glad to know that it's working. Glad to have been of assistance.

[Former Member · Accepted solution]

One thing I noticed while testing is it is pulling 70 characters but in the code it's substring (0, 70) right? that should be 71 chars or 70 chars?

thx

mike

[engswee · Accepted solution]

substring(0,70) will get 70 characters. It will take from beginIndex to (endIndex - 1), refer to the following JavaDoc description of substring function.

String (Java Platform SE 7 )

[Former Member · Accepted solution]

thanks....