Single flat file to multiple XML files

[Former Member]

Hi guys,

I'm trying to map a flat file into an IDOC. But I need to reformat the flat file before the mapping, where I need your help.

I will try to explain my problen with an example:

A) This my source flat file

Book1 000 0000 000

Account Date Amount Ref# (**this line won't be in the file, just FYI)

200 20090504 500 XYZ-123

100 20090504 400 XYZ-999

300 20090504 -400 XYZ-123

200 20090504 -400 XYZ-999

100 20090504 -100 XYZ-123

B) I ignore header and create a XML file with all lines

200 20090504 500 XYZ-123

100 20090504 400 XYZ-999

300 20090504 -400 XYZ-123

200 20090504 -400 XYZ-999

100 20090504 -100 XYZ-123

C) Now, I want to split the above XML file into individual files based on Ref#, so I would have:

File1:

200 20090504 500 XYZ-123

300 20090504 -400 XYZ-123

100 20090504 -100 XYZ-123

File2:

100 20090504 400 XYZ-999

200 20090504 -400 XYZ-999

How do I do that?

Edited by: Mohammad Islam on May 5, 2009 1:16 AM

[Former Member]

Hi Fatima, almost working

Current occurences:

Messages------------1..1
 Message1-----------1..1 	
  P1----------------0..unbounded
   C1---------------0..unbounded
    Account---------1..1
     Date ----------1..1
      Amount--------1..1
       Ref#---------1..1

Current Mapping:

Ref#(Context C1)--removeContext--sort(lex case-ins)--splitBYValue(value change)--collapseContext--P1

Ref#(Context C1)--removeContext--sort(lex case-ins)--splitBYValue(value change)--c1

Account(source Field)-----------Account(Target field)
Date(Source Field)--------------Date(Target field)
Amount(Source Field)------------Amount(Target field)
Ref#(Source Field) -------------Ref#(Target field)

Test source data:

Messages------------
 Message1----------- 	
  P1----------------
   C1---------------
    Account---------1
     Date ----------05122009
      Amount--------999
       Ref#---------111
   C2---------------
    Account---------2
     Date ----------05122009
      Amount--------699
       Ref#---------222
   C3---------------
    Account---------3
     Date ----------05122009
      Amount--------899
       Ref#---------111

Expected output:

Messages------------
 Message1----------- 	
  P1----------------
   C1---------------
    Account---------1
     Date ----------05122009
      Amount--------999
       Ref#---------111
   C3---------------
    Account---------3
     Date ----------05122009
      Amount--------899
       Ref#---------111
 P2----------------
   C2---------------
    Account---------2
     Date ----------05122009
      Amount--------699
       Ref#---------222

Actual output:

Messages------------
 Message1----------- 	
  P1----------------
   C1---------------
    Account---------1
     Date ----------05122009
      Amount--------999
       Ref#---------111
   C2---------------
    Account---------2
     Date ----------05122009
      Amount--------699
       Ref#---------222
  P2----------------
   C3---------------
    Account---------3
     Date ----------05122009
      Amount--------899
       Ref#---------111

What am I doing wrong?

[Former Member]

Did you map acount and all other fields also as below:

Account---->removeContext-->sortByKey---->Account

[Former Member]

Hi,

Map Like below.

Ref#(Changecontextto P1)sort-SplitByValue(ValueChange)CollapsecontextP1

Ref#(Changecontextto P1)sort-SplitByValue(ValueChange)--C1

1. Ref#(Changecontextto P1)sort-SplitByValue(eachvalue)--Ref#

2. Ref#(Changecontextto P1)--

---

SortByKey-SplitByValue(eachvalue)--Account

Account(Changecontextto P1)--

according to 2 map all the other fields.

Regards,

Prakasu.M

Edited by: prakasu on May 7, 2009 10:49 AM

Edited by: prakasu on May 7, 2009 11:07 AM

[jyothi_anagani]

Hi Islam,

I gave mappin above. Do like that you have to use sortByKey . I gave already check it and do.

Why you are getting this output is you are sorting only reference number and you are not sorting the Account and the remaining those. So use sortByKey . You will get it. Refer my previous post.

Thanks.

[Former Member]

Book1 000 0000 000

Source structure and Target structure same

Book1------------------0...unbounded
   P1--------------------1...unbounded
     Account  --------1
     Date -------------1
     Amount-----------1
     Ref#  --------------1

Use the mapping like given below and I have tested it .

ref#----removeContext-------sort------splitBYValue---------collapseContext------Book1
                                        (value change)

ref#----removeContext-------sort------splitBYValue---------P1
                                        (value change)

Account(source Field)---------------Account(Target field)
Date(Source Field) --------------Date(Target field)
ref#(Source Field) --------------ref#(Target field)

[Former Member]

Hi,

Follow the below mentioned logic, this may solve your problem.

use input filed->removecontexts>sortbykey->splitbyvalue(valuechanges)>collapsecontexts

for the highestnode(i.e parent node) in the output message type.

and use input filed->removecontexts>sortbykey--->splitbyvalue(valuechanges)

for all the Individual Fields in the output message type.

Regards,

Ramana.

[Former Member]

Hi,

May be I need to explain the problem better-

Source message has 1 parent node and 5 child nodes

p1--

c1)

200

20090504

500

XYZ-123

c2)

100

20090504

400

XYZ-999

c3)

300

20090504

-400

XYZ-123

c4)

200

20090504

-400

XYZ-999

c5)

100

20090504

-100

XYZ-123

I want to-

First, group my child nodes based on ref# (e.g. XYZ-123) and then create a parent node with the chld nodes for each unique ref#.

So for the above example, I should get 2 parent nodes:

p1--

c1)

200

20090504

500

XYZ-123

c2)

300

20090504

-400

XYZ-123

c3)

100

20090504

-100

XYZ-123

p2--

c1)

100

20090504

400

XYZ-999

c2)

200

20090504

-400

XYZ-999

[Former Member]

Hi,

Did u tried what i told u...

if u would have tried for parent node..u will get 2 files...i m sure for tht..(i hv edited one thing in previous reply)

plz check it once and let me know if any problem

Regards,

Manisha

[jyothi_anagani]

Hi Islam,

This is Your Source and Target Structure.

Book1------------------0...unbounded
   P1--------------------1...unbounded
     Account  --------1
     Date -------------1
     Amount-----------1
     Ref#  --------------1

Here sortByKey function is very Important

   Account------>removeContext------>sortByKey------>Account
                                         Ref#  /

Date--------------->removeContext------>sortByKey------>Date
                                         Ref#  /

Amount--------->removeContext------>sortByKey------>Amount
                                         Ref#  /

Ref#----------->removeContext-------->sort-------->Refe#

Ref#----------->removeContext-------->sort-------->splitByValue(Value Change)------>P1

DO LIKE THIS YOU WILL GET THE REQUIRED OUTPUT.

Thanks.

[Former Member]

Hi,

1...if u need to create number of files at target site then u need to change occurance of target to 1 to unbound.

2...Now it is not required that ur ref* field value come into sorted order so here u need to sort it first before performing any action. So it will be

Ref#---Remove context -sort-spilt by value (value change) -collapsecontext-target root node

This will create number of files depends on Ref# variation.

3...Now u need to map other fields so that it will go into proper file...so take care while mapping..

for each field to map...u need to do following

A) any source field removecontextsortbykey----

---

.....Ref#--

---

removecontext

B) Ref#(change context to root node)-sortspiltbyvalue(eachvalue)--

Map A and B like this

Formatyexample---splitbyvalue(valuechange)-collpasecontext-targetfield.

There may be few correction need to do..as i have not tried this now..but done long before

Regards,

Manisha

Edited by: Manisha Dahatonde on May 6, 2009 10:48 AM

[former_member581827]

Hi,

You can use source field Ref #->remove context-> SplitByValue(Change)-->create target node

Regards,

Chandra

[Former Member]

Hi, Mu:

I am assuming you have four col

Coloumn Possible Value

Col1 - 200, 100...

Col2 - 20090504,

Col3 - 500, 400, -400 ...

Col4 - XYZ-123, XYZ-999

I am also assuming you have done content conversion.

You can design your mapping as multi-mapping, because you want two messages in your target side:

Your source message type and target message type is same, in message tab of you mapping design,

check the target message occurance to 0:unbounded.

Your message Type node mapping:

Col4 -> RemoveContext -> Sort -> SplitByValue[valueChange]->CollapsContext->Your target message type

Other nodes mapping: take the Col1 as example:

Col1 -> removeContext -> SortByKey ------------------> FormatByExample -> Col1
Col4 -> RemoveContext --/                              /
Col4 -> RemoveContext->Sort->SplitByValue[ValueChange]/

Regards.

Liang

[Shabarish_Nair]

you will have to do this in your mapping.

use the sort and sortbykey functions to achieve this.

[former_member732072]

Hi Mohammed,

Please have a look at the following links and see if it helps you.

Best Regards