Schema question for a flat file parsing

Guest · August 25, 2004, 8:53am

Hello all:
Trying to get some direction on this:

I am having a problem creating schema for the following TICKET_RECORD flat file:
H 0 VALIDATION
Travelwithm Vacations 082520042004000242
Travelwitnm Vacations 082520042004000418248
Travelwithm Vtions
T 0 VALIDATION

SCHEMA DEFNITION (added space in front of segment id for H and T records

<SEGMENT ID="H ">
<POSITION NAME=“RECORD_TYPE” START=“0” LENGTH=“1”/>
<POSITION NAME=“FILLER” START=“1” LENGTH=“12”/>
<POSITION NAME=“SECURITY_QUALIFIER” START=“13” LENGTH=“1”/>
<POSITION NAME=“SECURITY_PASSWORD” START=“14” LENGTH=“10”/>
<POSITION NAME=“VENDOR_ID” START=“24” LENGTH=“12”/>
<POSITION NAME=“INTERCHANGE_DATE” START=“36” LENGTH=“6”/>
<POSITION NAME=“INTERCHANGE_TIME” START=“42” LENGTH=“4”/>
<POSITION NAME=“TRANSACTION_SET_TYPE” START=“46” LENGTH=“1”/>
<POSITION NAME=“CONFIRMATIONS_REQUESTED” START=“47” LENGTH=“1”/>
<POSITION NAME=“TEST_INDICATOR” START=“48” LENGTH=“1”/>
<POSITION NAME=“NUMBER_OF_RECORDS” START=“49” LENGTH=“5”/>
</SEGMENT>
<SEGMENT ID=“recordWithNoID”>
<POSITION NAME=“WHOLESALER_NAME_CHARACTER” START=“0” LENGTH=“50”/>
<POSITION NAME=“ARRIVAL_DATE” START=“50” LENGTH=“8”/>
<POSITION NAME=“RESERVATION_NUMBER” START=“58” LENGTH=“15”/>
<POSITION NAME=“LAST_NAME” START=“73” LENGTH=“15”/>
<POSITION NAME=“FIRST_NAME” START=“88” LENGTH=“15”/>
<POSITION NAME=“TICKET_CODE” START=“103” LENGTH=“5”/>
<POSITION NAME=“TICKET_DESCRIPTION” START=“108” LENGTH=“10”/>
<POSITION NAME=“TICKET_QUANTITY” START=“118” LENGTH=“3”/>
<POSITION NAME=“TICKET_COST” START=“121” LENGTH=“8”/>
<POSITION NAME=“HOTEL_CODE” START=“129” LENGTH=“30”/>
</SEGMENT>
<SEGMENT ID="T ">
<POSITION NAME=“RECORD_TYPE” START=“0” LENGTH=“1”/>
<POSITION NAME=“FILLER” START=“1” LENGTH=“12”/>
<POSITION NAME=“VENDOR_ID” START=“13” LENGTH=“12”/>
<POSITION NAME=“INTERCHANGE_DATE” START=“25” LENGTH=“6”/>
<POSITION NAME=“INTERCHANGE_TIME” START=“31” LENGTH=“4”/>
<POSITION NAME=“CONFIRMATIONS_REQUESTED” START=“35” LENGTH=“1”/>
<POSITION NAME=“TEST_INDICATOR” START=“36” LENGTH=“1”/>
<POSITION NAME=“NUMBER_OF_RECORDS” START=“37” LENGTH=“5”/>
</SEGMENT>
<SEG-ID>H</SEG-ID>
<SEG-ID>T</SEG-ID>
<LOOP Start=“recordWithNoID”>
<SEG

Chris_L · August 25, 2004, 7:00pm

Harsh,

Please don’t double post. Especially one such as detailed as yours.
Have your tried setting skipWhiteSpace value to false in convertToValues. Just a thought as I’m unsure of the output.

gupta_r.17495 · August 25, 2004, 7:05pm

Chris,

As far i know Harsh is not using WmEDI/convertToValues, he is using some custom made service for parsing.

And i didnt mentioned in the other post of using the WmEDI services but no details posted about that by him.

May be he can try to attempt of using this service,if he has WmEDI module installed over there.

Harsh,please try to maintain the posts which you have created earlier.

Regards,

gupta_r.17495 · August 25, 2004, 7:06pm

Chris,

As far i know Harsh is not using WmEDI/convertToValues, he is using some custom made service for parsing.

And i did mentioned in the other post of using the WmEDI services but no details posted about that by him.

May be he can try to attempt of using this service,if he has WmEDI module installed over there.

Harsh,please try to maintain the posts which you have created earlier.

Regards,

Guest · August 25, 2004, 7:55pm

Sorry about the double post Chris:
I will try the skipWhiteSpace hint.

Hello RMG:
Yes I am using the convertToValues service within:
here is my flow:
getTemplate
convertToValues.

The problem comes in my record structure which has TICKET_RECORD/H(space) or TICKET_RECORD/T(space), the TICKET_RECORD/recordwithNoID values are populated perfectly from the [ouput] of the convertToValues service, and this is because in my schema i had to put segment ID for H and T record with a space, to pull up the appropriate record from the flat file.

Appreciate the response, thanks

harsh

Guest · August 25, 2004, 8:12pm

RMG: i am attaching my flow service and flat file with schema i am running thru this, i tried Chris, suggestion also but i still have the problem.

Thanks all!
Harsh

Scema and flat files
schema_flatfile.txt (5.9 k)

test.zip (4.5 k)

gupta_r.17495 · August 25, 2004, 8:36pm

Harsh,

If you handling the problem with Space and recordWithNoID loop is working then what is the issue you have?

you mean when data is there in H/T records then it should overwrite the spaces right and still recordWithNoID loops should work.

Chris_L · August 25, 2004, 8:52pm

RMG,

I think the problem he’s having is with the BRANCH. The schema is fine but he can not BRANCH on "record/H " only on “record/H”.

Guest · August 25, 2004, 9:10pm

Hello Chris/RMG:
Thanks for the reply, Yes the problem i am having is when i branch as i now have a space added in the [output] record . "record/H ". So i cannot proceed further. Is there somewhere i can avoid this space and still parse the file properly.
regards and thanks

harsh

Chris_L · August 25, 2004, 9:46pm

try trimming the recordid field to make it “record/H” and then BRANCH on it.

Guest · August 25, 2004, 10:22pm

Chris: can you explain in some detail on how i can do that.
Thanks!

harsh

Chris_L · August 25, 2004, 10:56pm

Use a map statement with trim as a transformer.
record.recordId ---- trim — record.recordId
should result in
"H " —> “H”
Just map the record back to itself using trim in the transformer step.
Then you should be able to BRANCH on record.recordId=“H”

HTH

gupta_r.17495 · August 25, 2004, 11:29pm

Harsh,

Chris sugesstion should resolve your problem.
please try to implement the same way he mentioned.

Guest · August 26, 2004, 12:12am

Appreciate the response Chris/RMG but this still doesnt solve the problem:
This is how my TICKET_RECORD looks like:

converToValues[OUTPUT]–
maps—>
TICKET_RECORD
|

H<space>(HEADER record)
-string values
recordWithNoID (repeating record)
-string values
T<space>(TRAILER RECORD)
-string values

The Header gets mapped from output into a H(space), i am not sure how i can use a trim , string transformer to trim a record.
My branch is for values under H ( header) or T (trailer):
TICKET_RECORD/H

regrds,
Harsh

gupta_r.17495 · August 26, 2004, 12:32am

Can you use two different Branch steps one having with H(Space) and the other with no space and similarly for T Record.

Will this work for you?

sorry let me also think if anyother options.

Chris_L · August 26, 2004, 12:35am

You’re correct but aren’t you BRANCHing on the field withing the record.
ie. TICKET_RECORD/H /field

Guest · August 26, 2004, 1:05am

RMG: In tried giving TICKET_RECORD/H(SPACE), but this space is not persisted in developer and disappears when i save the change in the branch.

Chris i finally ended up using Record/H /field to go into the desired sequence.

Going back to the schema is there a more effcient way to tackle this, i.e records with no id’s having values similar to Header and Trailer segment id?.

Thanks all for your timely suggestions and ideas, i am going to the next level :-).

Harsh

gupta_r.17495 · August 26, 2004, 3:08am

This is also a good option using Record/H /field to go into the desired sequence.Anyways you got the resolution.

goodluck,

Guest · April 11, 2005, 12:08pm

hi,

I am using convertToValues for a fixed length flat file.
My input flat file has child records which can appear in any order.
for example:

MainRecord…
childrec1…
childrec1…
childrec2…
childrec2…
childrec1…
childrec1…

i have set ordered=false in my schema. But in the output i am gettting 2 separate doc list for childrec1. I want a single separate doc lists for childrec1 and childrec2. Has anybody faced this problem. Please give me solution for this problem of mine.
thanks,
naka

gupta_r.17495 · April 11, 2005, 7:02pm

Noor,

Once you have parsed the flatfile you may have to map the seperated doclists what ever childrec1,childrec2 to another single doclists(childrec1),otherwise flatfile schema cannot handle this type of requirement since childrec1 is coming under childrec2.

May be there is easy way,let c if anyone have another clue to resolve.

HTH,
RMG.