Breaking text file into multiple IDOC

[Former Member]

Hi all Expert

I am using file (TEXT) to IDOC (Wpuums01) scenario—POS integration

In one text file number of line item around 3000 to 5000 so one IDOC is generated for one text file in XI side that contain around 7500 sgement to 9000 segments

I want maximum 1000 line item will come in one IDOC if number of item greater then 1000 in text file then another IDOC of same store is generated is it possible

Thanks

Amit Shivhare

Edited by: Amit Shivhare on Mar 14, 2008 8:33 PM

Edited by: Amit Shivhare on Mar 14, 2008 8:36 PM

[Former Member · Accepted solution]

If one line can be considered as one recordset, in file sender adapter, you can specify max number recordsets per message, to explore if that feature solve your problem.

Regards

Liang

[Former Member]

1. Ensure that all the lines of the file come into one Message through FCC.

2. Change the occurence of the IDOC node to 1..UNBOUNDED

3. While mapping from source node to the IDOC node, write an UDF to check the condition and determine how many instances of IDOC nodes have to be generated ( for example two instances if the number of line items in source are more than 10000),

4. Generate the instances of line items of IDOC using condition.

It is basically about generating required number of instances of target structures through conditions in UDF

Edited by: Ramesh on Mar 16, 2008 12:47 AM

[nisarkhan_n]

Recordset per message entry will pick the number of records mentioed here and make it as a single XI message and process it...

suppose your file has 9000 records you can mention the recordset per message as 1000 then it will create 9 messages and each message genereates one IDOC.....

this method is good only when you dnt have any dependency in reocrds in text file like Header and line item...if that is the case then this is not right option.

[shivhare]

Hi Expert

My text file structure is

1#C3025#20080128

2#000000040000000448##-# 2## 990#NONRETG#-#ZPR1# 297#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 26.66

2#000000040000000449##-# 1## 595#NONRETG#-#ZPR1# 261.8#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 12.82

2#000000040000000450##-# 1## 995#NONRETG#-#ZPR1# 398#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 22.96

2#000000040000000451##-# 1## 1095#NONRETG#-#ZPR1# 443.47#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 25.06

2#000000040000000452##-# 5## 5575#NONRETG#-#ZPR1# 1350#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 162.5

2#000000040000000453##-# 2## 1190#NONRETG#-#ZPR1# 594.94#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 22.88

2#000000040000000454##-# 2## 1390#NONRETG#-#ZPR1# 695#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 26.73

2#000000040000000455##-# 2## 140#NONRETG#-#ZPR1# 42#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 0

2#000000040000000456##-# 1## 220#NONRETG#-#ZPR1# 66#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 5.92

2#000000040000000457##-# 1## 495#NONRETG#-#ZPR1# 99#-#ZPR2# 0#-#ZPR3# 0##ZVAT# 15.23

first line contain header data with respect to that header data We have find out store code and posting date

If I am not wrong when we have generated multiple IDOC of one main text file then we also required to repeat header data as per after required line item so how can do it help me

Thanks

Amit Shivhare

[Former Member]

Hi Amit

I am having the same problem as you, one header line and 10,000 transactions, and I have to map the same header for every IDOC I generate. My XI can process only small files (1 header and 100 transactions).

Did you figure out how to split your huge message ?

[Former Member]

Hi,

You can try the option if you are using FCC in sender file adapter, use recordset per messages=1000

OR

Use this Logic

/people/michal.krawczyk2/blog/2005/12/04/xi-idoc-bundling--the-trick-with-the-occurance-change

Regards

Seshagiri

[Former Member]

As Liang said, In the File Sender Adapter, under FCC parameter "recordSetPerMessage" define it as 1000 orwhat ever number you want to process