Originally posted by: SystemAdmin

Can I have multiple input cards with multiple occurrences for a functional map? Below is the rule that I'm using:
=MyVehicleFunction(VehicleRecord:FSC_File,LOOKUP(LienholderRecord:FSC_File,RecordType Lienholder:LienholderRecord:FSC_File="11"))
What it does is: The LOOKUP function just takes the first Lienholder record and appends to all the Vehicle records. The way it should work is 1st Lienholder record should append to 1st VehicleRecord and 2nd Lienholder record should append to the 2nd VehicleRecord and so forth. If I've only one Lienholder record and 2 vehicle records, the Lienholder record should append to the 1st vehicle record and the 2nd vehicle record will not have any Lienholder information. How do I modify the LOOKUP function to work for multiple occurrences? Or is there any other function that I could use? I tried to use EXTRACT function, but it produces Cartesian product(doulbe the records). Thanks a lot!!!

Message was edited by: rathna
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

"What it does is: The LOOKUP function just takes the first Lienholder record and appends to all the Vehicle records."

And that is what you're telling it to do. The number of vehiclerecord in the file is the number of output records you're getting, right? So what you're saying is run the functional map for each VehicleRecord, and for each run, bring over the first LIenholderrecord that has the record type as "11", and use that same record every time.
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

Both of these records (VehicleRecord:FSC_File,LienholderRecord:FSC_File) are in the same file?

If so, what does your input type tree look like, i.e.

File,
...GroupOfRecords(1:s)
.......VehilceRecord(1:1)
.......LienholderRecord(1:5)
???
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

Yes, both of these records are in the same file.
File,
...GroupOfRecords
.......VehilceRecord(s)
.......LienholderRecord(s)
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

I'm assuming that the "groupofrecords" in your type is reoccurring, so your file looks like this inside, multiple groups of records, and each group contains at least one VehilceRecord, and 0 or more LienholderRecord :
VehilceRecord
LienholderRecord
VehilceRecord
LienholderRecord
VehilceRecord
VehilceRecord
VehilceRecord
LienholderRecord
???

What version of Datastage are you using?
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

Yes, each group contains atleast one VehicleRecord and 0 or more LienholderRecord.
VehicleRecord
VehicleRecord
LienholderRecord
LienholderRecord

OR

VehicleRecord
VehicleRecord
LienholderRecord

OR

VehicleRecord
VehicleRecord

I'm using WTX 8.1.0.2
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

Look at the attached, but what you're saying is fundementally wrong.

Just say you have a "group" of records containing one VehicleRecord with no LienholderRecord, then another "group" of record containing two VehicleRecord and one LienholderRecord, it will "look" like :
VehicleRecord (group 1)
VehicleRecord (group 2)
VehicleRecord (group 2)
LienholderRecord (group 2)

The map will see this as one record group, and by the logic you described, vehicleRecord (group 2) goes with LienholderRecord (group 2), but the map will think VehicleRecord (group 1) and lienholderrecord( group 2) go together.

If you have only one group per file, the attached code will work.
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

Is there field in the vehicle recs to distinguish the groups from each other? Like, is there a vehicle number field that is different for each group of vehicle records?
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

Do you have any email id that I can send my type trees and the map?
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

Could you not download the attached maps/type trees? They are in 8.1. (build 114).
You understand what I mean on how it needs to differentiate the different groups?

Can you zip and upload stuff here?

Message was edited by: rep
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

Here are the attached files.
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

First, you have "record type" for identifying your records, but you're using required fields. I.E, If that field is not there, then validate the current record against the next record defintion. I would use a component rule. Open up each record in the type tree editer. Right click the "record type" field and select "identifier", just to be on the safe side. In the rule next to each record type field, put $="X", where X = the record type for that record. (applicant = 1, driver record = 2, etc). This way, it doesn't have to validate all the fields until it gets to that one field that's missing, it looks at the first field and knows what record it is. This will save resources. For example, run your map now and look at the size of the trace file, then put this change in, then look at the size.

Now check out the rule in the map I sent you in the F_mapGroup functinoal map.

The EITHER function says "do the first thing that does not return "NONE".

What I say is EITHER do a functional map where the vehiclerecord is present and where there is a lienholder record with the same INDEX in the file, OR, do a functinal map where there's just a vehicle record.

When you say F_MAP(vehiclerecord, lookup(leinholderrecord, field=field)), and there is no leinholderrecord, it returns NONE, so you will not get output for each vehiclerecord that doesn't have a lienholder record.

I see you did a IF(PRESENT(leinholder), Myvehicle(vehicle, lienholder, else, Myvehicle(vehicle, " "), but you can't do that. The type tree in the F_map has to match the data you're passing it.

You may have to code another F_map with just the vehicle record and do the EITHER like I did, but it should work.

Message was edited by: rep
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

So by adding identifier to the "record type" field, even that record is missing in the input file, it will still continue to process the file right? Or how do I continue through the rest of the file, if some of the records missing(since they are not mandatory)in the input file?

I've added one more F_Map with just VehicleRecord, and used EITHER like how you did, and it worked. I really appreciate your help. Thanks a lot...

So I can use the same strategy for other records for example, CoverageUnderwriting is also part of the VehicleRecord. I should create one more F_Map for the CoverageUnderwriting record and include in the rule right? Is it a good idea or is there any other solution that I could use?

Thanks...
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

"So by adding identifier to the "record type" field, even that record is missing in the input file, it will still continue to process the file right?"

Just like any of the fields (or any other object), if a record (just another object) is not mandatory (S), it's not mandatory. If it's not mandatory, it being absent is valid. In the event you ever have a file that has multiple transactions, like just say the input you sent me had mulitple applicant and related records many times over in the file, and one of the mandatory records was missing, the entire file would not process, unless, you put the "restart" attribute on the highest group level, which says, "if any object (field or record) is missing a mandatory field, or any field is invald (text field in a numeric field)", fail this one transaction, but continue processing the rest of the file.
"I've added one more F_Map with just VehicleRecord, and used EITHER like how you did, and it worked. I really appreciate your help. Thanks a lot..."

I'll let you know where to send the check. Actually, can you get me a discount on my insurance??? ;)

"So I can use the same strategy for other records for example, CoverageUnderwriting is also part of the VehicleRecord. I should create one more F_Map for the CoverageUnderwriting record and include in the rule right? Is it a good idea or is there any other solution that I could use?"

I don't see a "CoverageUnderwriting" in either type tree (input or output). Is it on the input?

If so.....
Anytime you're going to have different defintions, or more input cards for a functional map, you'll need a new one. (like in the case we just worked, you needed one functional map with one input, and one functinal map with two inputs. Let me know a bit more detail on what you mean with the coverageUnderwriting.
There has never been a situation in mapping that I could not handle with Datastage, and there has rarley been a situation that couldn't not be hanlded in more then one way. Some of it could get complicated and you may need more experince before I start spitting out complex coding possiblities, such as partitioning as a solution.
#IBM-Websphere-Transformation-Extender
#DataExchange
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

I'll let you know where to send the check. Actually, can you get me a discount on my insurance??? ;)
Sure, If you want to pay more... ;-)

Input file has a record 16 for each vehicle category and each coverage. I've to map each Vehicle and coverage record from typetree(QuoteCovDiscountSurchargeRecord) to a different elements in the Vehicles\Vehicle\Coverages\ output file.
For example: Vehicles\Vehicle\Coverages\BI, from input file for the first vehicle the record looks like this:
"16","17","1","1","Cov","V","1","BI","BI","Bodily Injury","SplitKDlr","25/50","141.00","","","","","","","","","","","","Y"

The identifiers we could use here are 1st(RecordType),7th(Iteration) and 8th(TypeofCoverage) positions from the QuoteCovDiscountSurchargeRecord input typetree.
But how do I iterate through each 16th record to map to each coverages in the Vehicles\Vehicle\Coverages\ output file?
What rule I should create for the Vehicle element instead of
EITHER(MyVehicleAddIntFunction (VehicleRecord:FSC_File,EXTRACT(LienholderRecord:FSC_File,INDEX(VehicleRecord:FSC_File)=INDEX(LienholderRecord:FSC_File))),
MyVehicleFunction(VehicleRecord:FSC_File))

Message was edited by: rathna
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

Now I'm really confused.

So in MyVehicleAddlFuctnion and in MyVehcileFunction, you also need information from the record with the record type = 16, the QuoteCovDiscountSurchargeRecord??
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

I'm really sorry for the confusion.
Yes, I need information from the QuoteCovDiscountSurchargeRecord with the record type = 16 in MyVehicleAddlFunction and MyVehicleFunction.
Apart from QuoteCovDiscountSurchargeRecord, I've some other records, that I'll have to extract information from those records and include in the MyVehicleAddlFunction and MyVehicleFunction.

Right now I'm concentrating on 16th record, since for each vehicle, each coverage, I've separate 16th record. How do I loop through each 16th record, each vehicle and each coverage and map to separate elements in the Vehicles\Vehicle\Coverages\?

For example:
"16","17","1","1","Cov","V","1","BI","BI","Bodily Injury","SplitKDlr","25/50","141.00","","","","","","","","","","","","Y"
"16","17","2","1","Cov","V","2","BI","BI","Bodily Injury","SplitKDlr","25/50","114.00","","","","","","","","","","","","N"

The above records will be mapped to two different vehicles BI coverage(Vehicles\Vehicle\Coverages\BI).

"16","17","3","1","Cov","V","1","PD","PD","Property Damage","KDlr","25","105.00","","","","","","","","","","","","Y"
"16","17","4","1","Cov","V","2","PD","PD","Property Damage","KDlr","25","84.00","","","","","","","","","","","","N"

The above records will be mapped to two different vehicles PD coverage(Vehicles\Vehicle\Coverages\PD).

Hope the above explanation make sense...Thanks a bunch!!!

Message was edited by: rathna
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

Let me give some information, and I hope it helps.

In the event there is a way to grab the information out of the the "16" record based on the information in the vehicle record or lienholden record, and there is a one on one relationship, where for each 3 record (vehicle) , you need information from one 16 record (QuoteCovDiscount).....

First, you need to add the entire input file to each of the functional maps MyVehicleAddlFuctnion and in MyVehcileFunction.

It will look like this when you're done.

EITHER(
MyVehicleAddIntFunction(VehicleRecord:FSC_File,
EXTRACT(LienholderRecord:FSC_File,INDEX(VehicleRecord:FSC_File)=INDEX(LienholderRecord:FSC_File)),FSC_File),
MyVehicleFunction(VehicleRecord:FSC_File, FSC_File ))
In order to do this, you have two options; one being a much better way. The bad way for the inexperinced, is to delete each of the functinal maps, add the input file to the f_map's, and rebuild them from scratch, or....

Using the MyVehicleFunction map first, add the FSC_FILE to the functinal map rule:

MyVehicleFunction(VehicleRecord:FSC_File, FSC_File )
Go to the map rule and add TEST for the F_map name (or whatever extra character you like):

MyVehicleFunctionTEST(VehicleRecord:FSC_File, FSC_File )

Now build the new f_map using the f_map wizard.

You now have two functional maps, MyVehicleFunction and MyVehicleFunctionTEST.

Click on "list" mode in the map navigator, instead of composition. You will see both maps. Open them up, then open up input cards. Hold down the "ctrl" key and drag the second input card from MyVehicleFunctionTEST to MyVehicleFunction.

Go back to the map rule and remove "TEST" from the f_map name. Now the original MyVehicleFunction has the entire input file as an input card. Do the same for MyVehicleAddIntFunction.

Looking at this:

"16","17","1","1","Cov","V","1","BI","BI","Bodily Injury","SplitKDlr","25/50","141.00","","","","","","","","","","","","Y"
"16","17","2","1","Cov","V","2","BI","BI","Bodily Injury","SplitKDlr","25/50","114.00","","","","","","","","","","","","N"

The above records will be mapped to two different vehicles BI coverage(Vehicles\Vehicle\Coverages\BI).
Do mean those two records will be mapped like:
Vehicles
Vehicle
BI1
BI2

Or

Vehicles
Vehicle1
BI1
Vehicle1
BI2

????
I'm starting to get very busy at work, so what I can offer may quickly be reduced.

Message was edited by: rep
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

The two records will be mapped like below:

Vehicles
Vehicle1
BI ("16","17","1","1","Cov","V","1","BI","BI","BodilyInjury","SplitKDlr","25/50","141.00","","","","","","","","","","","","Y")
Vehicle2
BI("16","17","2","1","Cov","V","2","BI","BI","Bodily Injury","SplitKDlr","25/50","114.00","","","","","","","","","","","","N"
)
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

I've added that entire input file(FSC_File) to MyVehicleAddlFuctnion and MyVehcileFunction maps as you suggested.
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

Hold up. What I said was:

In the event there is a way to grab the information out of the the "16" record based on the information in the vehicle record or lienholden record, and there is a one on one relationship, where for each 3 record (vehicle) , you need information from one 16 record

From what I'm getting from your last post, that is not the way this will be mapped. Eitherway, if you didn't know how to put the extra input card in like that, it is good practice. You will undoubtably need to add an input card in the future, and the only other way is much more work.

But this quesiton is very important:
So the number of vehicle/vehicles in the output is not going to be based on the vehicle records (regardless of wether they have a lienholder rec), it's going to be based on the number of 16 records?? There is one one BI record on the output under vehicle, so if have two BI records that match one vehicle, you need two output vehicles loops, each with the same information out of the vehicle and lienholder record, the first one will have the first BI, then the second BI, and if there's a PD 16 type record, it's going to be the same vehicle infomation, but the PD record is going to be populated. Is that right?

And now that I think about it, you have 2 BI records and 3 PD records for the same vehicle, right?

So it's it going to be:

vehicle
vehicles1
BI1
PD1
vehicles1
BI2
PD2

???
Actually, it's it going to be

vehicle
vehicles1
BI1
PD1
UMBI1
UMPD1
vehicles1
BI2
PD2
TOWING2
...................

Your answer may change everything.
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

No, acutually for each vehicle I'll have one BI, one PD, one UMBI, one UMPD etc.,

Vehicles
..Vehicle1
...Coverages
....BI
....PD
....UMBI
....UMPD
.
.
.
.../Coverages
../Vehicle1
..Vehicle2
...Coverages
....BI
....PD
....UMBI
....UMPD
.
.
.
.../Coverages
../Vehicle2
Vehicles

We can identify for which vehicle(either for 1st or 2nd or 3rd vehicle), this 16th record belongs to is by looking at the 7th field(Iteration)of the 16th record. And the 8th field is for the type of coverage(BI or PD or UMBI or UMPD etc.,) So that's why I'll have many 16th records for each vehicle, each coverages. So for every 3rd record that is for each vehicle, I've to loop through 16th records to get all the coverages(BI, PD, UMBI, UMPD etc) for that vehicle and map to the output and continue the same for all the other 3rd records.

Message was edited by: rathna

Message was edited by: rathna
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

I hope I'm getting what you're saying now.

So putting the entire file as an input card to the fmaps was the right thing to do.

So, now you have the entire file of data at your disposel while in the f_maps.

On the BI field, it's

=EXTRACT(16th record data, 7th iteration of 16th record = matching field in vehicle record & 8th field = "BI")

Same thing for the PD info, just replace BI with PD, and so on. In the event you have to map the entire 16 record into the field, use the TEXT function.

If there is only one object on the input (16th record BI coverage) and only one output record (Coverages/BI), there is no need for a loop. Just drag the info over.
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

Sorry to bother you...
I'm getting "Output argument of rule does not match output item sub-class" error message after adding the rule to the BI field for both the f_maps. What am I doing wrong here? Thank you so much for your help....

M115 ERROR:
Map: MyVehicleAddIntFunction Output: BI Comp:ElemDecl Coverages:XML_Vehicle
Output argument of rule does not match output item sub-class:
=EXTRACT(QuoteCovDiscountSurchargeRecord:FSC_InputFile,IterationType QuoteCovDiscountSurcharge:QuoteCovDiscountSurchargeRecord:FSC_InputFile=Occurrence Vehicle:FSC_Vehicle
& FSCKeyword QuoteCovDiscountSurcharge:QuoteCovDiscountSurchargeRecord:FSC_InputFile="BI")

M115 ERROR:
Map: MyVehicleFunction Output: BI Comp:ElemDecl Coverages:XML_Veh
Output argument of rule does not match output item sub-class:
=EXTRACT(QuoteCovDiscountSurchargeRecord:FSC_InputFile,IterationType QuoteCovDiscountSurcharge:QuoteCovDiscountSurchargeRecord:FSC_InputFile=Occurrence Vehicle:FSC_Veh
& FSCKeyword QuoteCovDiscountSurcharge:QuoteCovDiscountSurchargeRecord:FSC_InputFile="BI")

Message was edited by: rathna
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: janhess

The field your rule is in doesn't match th data type that the rule is returning - ie group receiving an item, item receiving a series.
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

"In the event you have to map the entire 16 record into the field, use the TEXT function."
If you're taking one field from the BI record and putting it into the BI on the output, it'll work. If you're putting the entire input record into the output record, you need to use the TEXT function.
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

Thanks a lot for all the help....

Message was edited by: rathna
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

Thanks a lot for your help... I made the changes according to your suggestion, and it is working now...
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender