I have a file :
FileType00000001~CustomerLoad~83e3ec9b-d824-48a5-8102-773b72ad8cca~20120717101648
#KeyDefinitionField~ImpGuid~RowId~FirstName~LastName~RsaId~PassportNumber~IfaNumber~HomeTelephone~WorkTelephone~MobileTelephone~Fax~EmailAddress
#KeyDefinitionField~ImpGuid~RowId~PolicyNumber~PolicyStatus~PolicyPremium
Customer00000001~~0~Nobongile~Langa~~~~~~~~
Policy0000000001~~0~402060005~I~99.00
Customer00000001~~1~Nobongile~Langa~~~~~~~~
Policy0000000001~~1~610473373~I~135.00
Customer00000001~~2~Nobongile~Mabunda~~~~~~~~
Policy0000000001~~2~403897536~I~99.00
Customer00000001~~3~Nobonile~August~~~~~~~~
I want to be ale to read line by line, check whether it is a Customer00000001 or Policy0000000001 and populate 2 different tables depending on what row i am in .
apologies still new to talend (java) .
regards
okay ... that is what i was looking for. a final question ? the entire row is put in one column, must I write to a temp file and extract from that temp file into the DB (separating the tilde delimited values) or is there and easier way. I aso suppose i need to do this in 2 cycle , one for Customer0000001 and one for Policy00000001. Regards
Hi Yes. You need to create two subjobs for this. tFileInputDelimited--main-->tExtractRegexFields--main-->tExtractPostionalFields--main-->Output The tExtractPostionalFields can help you extract columns with fixed pattern. Regards, Pedro
Hi, You could also split the row by a tMap. Create 2 output rows from one tMap where you filter the output on the substring Customer and Policy this way you only need to parse your input file once. Regards, Arno
