tFileOutputMSDelimited -> can it exclude key columns in the output

[Anonymous]

Is there a way for the tFileOutputMSDelimited component to be configured in a way so that it excludes the key columns (used in matching with parent key columns) in the output.
see images attached.
1 - 6
Current Output
-----------------
OrderHeader,1,Warren
OrderDetailQty,1,1,wLf,92
OrderDetailQty,3,1,Z4U,38
OrderDetailQty,4,1,IrV,8
OrderDetailQty,7,1,OTo,98
OrderNotes,1,1,q2wwDthXWn
OrderNotes,3,1,ecZJtARKbB
OrderHeader,2,Ronald
OrderDetailQty,2,2,4Sw,17
OrderDetailQty,5,2,HKl,13
OrderDetailQty,6,2,jTh,39
OrderNotes,2,2,gq3cR7gC0b
Desired Output
-----------------
OrderHeader,1,Warren
OrderDetailQty,1,wLf,92
OrderDetailQty,3,Z4U,38
OrderDetailQty,4,IrV,8
OrderDetailQty,7,OTo,98
OrderNotes,1,q2wwDthXWn
OrderNotes,3,ecZJtARKbB
OrderHeader,2,Ronald
OrderDetailQty,2,4Sw,17
OrderDetailQty,5,HKl,13
OrderDetailQty,6,jTh,39
OrderNotes,2,gq3cR7gC0b

[Anonymous]

Hi
It is impossible to exclude the parent key by configuring the tFileOutputMSDemilted component, you need to read the current output file again, regards it as a temp file, parse the file and output the result to a final desired file, here I show you how to do it:
in.csv:
OrderHeader,1,Warren
OrderDetailQty,1,1,wLf,92
OrderDetailQty,3,1,Z4U,38
OrderDetailQty,4,1,IrV,8
OrderDetailQty,7,1,OTo,98
OrderNotes,1,1,q2wwDthXWn
OrderNotes,3,1,ecZJtARKbB
OrderHeader,2,Ronald
OrderDetailQty,2,2,4Sw,17
OrderDetailQty,5,2,HKl,13
OrderDetailQty,6,2,jTh,39
OrderNotes,2,2,gq3cR7gC0b
out.csv:
OrderHeader,1,Warren
OrderDetailQty,1,wLf,92
OrderDetailQty,3,Z4U,38
OrderDetailQty,4,IrV,8
OrderDetailQty,7,OTo,98
OrderNotes,1,q2wwDthXWn
OrderNotes,3,ecZJtARKbB
OrderHeader,2,Ronald
OrderDetailQty,2,4Sw,17
OrderDetailQty,5,HKl,13
OrderDetailQty,6,jTh,39
OrderNotes,2,gq3cR7gC0b
Best regards
Shong

[Anonymous]

Dear Shong: thanks for a quick response. I think it does exactly what I needed.
Your solution is a great workaround for a quite useful option which can be a part of the tFileOutputMSDelimited component itself.
In future, there could be an option or *feature* to choose the schema of the different output rows, thereby incorporating the filter column option within the component itself.
my 2 cents.
Pankul

[Anonymous]

Hi,
i am new to talend, could anyone help me about the below task,
My Input is :
<?xml version="1.0" encoding="UTF-8"?>
<COMPANY>
<COMPANY_DETAILS>
<COMPANY_HEADER>
<COMPANYNAME>IBM</COMPANYNAME>
<COMPANYADDRESS>HYDERABAD</COMPANYADDRESS>
<COMPANYPHONE>9988776655</COMPANYPHONE>
</COMPANY_HEADER>
<EMPLOYEES>
<EMPLOYEE>
<EMP_ID>101</EMP_ID>
<EMP_NAME>SAN</EMP_NAME>
<EMP_DESIGNATION>SW</EMP_DESIGNATION>
<EMP_SALARY>50000</EMP_SALARY>
</EMPLOYEE>
<EMPLOYEE>
<EMP_ID>102</EMP_ID>
<EMP_NAME>SHAN</EMP_NAME>
<EMP_DESIGNATION>SW</EMP_DESIGNATION>
<EMP_SALARY>20000</EMP_SALARY>
</EMPLOYEE>
<EMPLOYEE>
<EMP_ID>103</EMP_ID>
<EMP_NAME>SANTH</EMP_NAME>
<EMP_DESIGNATION>SW</EMP_DESIGNATION>
<EMP_SALARY>8000</EMP_SALARY>
</EMPLOYEE>
</EMPLOYEES>
</COMPANY_DETAILS>
<COMPANY_DETAILS>
<COMPANY_HEADER>
<COMPANYNAME>CAP</COMPANYNAME>
<COMPANYADDRESS>BANGALORE</COMPANYADDRESS>
<COMPANYPHONE>9988776655</COMPANYPHONE>
</COMPANY_HEADER>
<EMPLOYEES>
<EMPLOYEE>
<EMP_ID>104</EMP_ID>
<EMP_NAME>SHIV</EMP_NAME>
<EMP_DESIGNATION>SW</EMP_DESIGNATION>
<EMP_SALARY>10000</EMP_SALARY>
</EMPLOYEE>
<EMPLOYEE>
<EMP_ID>105</EMP_ID>
<EMP_NAME>SHAKE</EMP_NAME>
<EMP_DESIGNATION>SW</EMP_DESIGNATION>
<EMP_SALARY>8000</EMP_SALARY>
</EMPLOYEE>
</EMPLOYEES>
</COMPANY_DETAILS>
<COMPANY_DETAILS>
<COMPANY_HEADER>
<COMPANYNAME>IMBPVT</COMPANYNAME>
<COMPANYADDRESS>CHENNAI</COMPANYADDRESS>
<COMPANYPHONE>9908778965</COMPANYPHONE>
</COMPANY_HEADER>
<EMPLOYEES>
<EMPLOYEE>
<EMP_ID>107</EMP_ID>
<EMP_NAME>SYDN</EMP_NAME>
<EMP_DESIGNATION>SW</EMP_DESIGNATION>
<EMP_SALARY>9000</EMP_SALARY>
</EMPLOYEE>
</EMPLOYEES>
</COMPANY_DETAILS>
</COMPANY>

i am excepting the below Output :
IBM,HYDERABAD,9988776655
101,SAN,SW,50000
102,SHAN,SW,20000
103,SANTH,SW,8000
3
CAP,BANGALORE,9988776655
104,SHIV,SW,10000
105,SHAKE,SW,8000
2
IBM PVT,CHENNAI,9908778965
107,SYDN,SW,9000
1
3,2,1 are the total records.
Thanks,
RK

[Anonymous]

Hi rkalla
In order to get the expected output as you described, you need to store the COMPANY_HEADER data and EMPLOYEES data to memory or a temporary file first, and then iterate each COMPANY_HEADER data, I have created a demo job to achieve this:

IBM;HYDERABAD;9988776655
101;SAN;SW;50000
102;SHAN;SW;20000
103;SANTH;SW;8000
3
CAP;BANGALORE;9988776655
104;SHIV;SW;10000
105;SHAKE;SW;8000
2
IMBPVT;CHENNAI;9908778965
107;SYDN;SW;9000
1

if you would like to take a look at the demo job, please send me an email.
Shong