File structure validation in PLSQL

Question

Hello All,
I need few clarification about this if i want to validate the file structure how can i do in plsql, below is the example where my files looks like.
IIORRIDGE
VA1601800700028120000000008+0000123822092013
VA1601800700029040000000008+0000323822092013
VA1601800700030860000000008+0000123822092013
VA1601800767943220000000008+0000123822092013
end;
Please help me
Mahesh

Answer 1

I need few clarification about this if i want to validate the file structure how can i do in plsql
Foe example like this line
VA1601800700028120000000008+0000123822092013
There are totally 44 characters if its more than 44 i want to alert it and if it is less than 44 also i want  to alert it
Like ablve line there will be more than 100000 lines will be der.
The problem of 'validate the file structure' implies that there could be issues with the structure of the file or the records.
And therein lies the problem if you try to use an external table since a table, or external table, is generally expected to have a known number of columns of known datatypes.
So first you need to document the business rules that you want to implement:
1. What are ALL of the validation rules that you need to check for?
  a. what record delimiters are expected (e.g. LF, CRLF, etc)?
  b. do you need to validate that the proper delimiters have been used?
  c. how are the fields of a record delimited? Is it a delimited file (by commas or other?), fixed-width, etc.
  d. what is the field delimiter (if delimited)?
  e. what are the datatypes for each field of the record
  f. can fields be NULL? If so, how is a null field indicated?
  g. do all records have to have the same number of fields? Or is TRAILING NULLCOLS allowed?
2. How do you want to handle all of the errors that you detect?
  a. reject the entire file if there is an error in ANY record?
  b. reject only the records that have an error but accept the ones that do not have errors?
3. How do you plan to allow users access to the problem records?
  a. give them access to the actual BAD file containing the raw data?
  b. give them access to a DB table containing the raw data?
4. How do you plan to reprocess ONLY the bad data once it has been corrected?
  a. reload the entire file?
  b. reload a new file containing the fixed records?
5. How do you plan to allow users to access the validated, good data?
  a. give them access to a DB table?
Sometimes the 'best' solution is to NOT load data into Oracle that isn't already validated.
You could write a standalone Java application to perform this 'file validation' and either produce 'clean' files that you need load into Oracle or the app could validate a file and load the clean data at the same time.
But until you document ALL of the requirements in terms of handling the 'dirty data' it is too early to be talking about the 'best' solution.