Check out the new video on my YouTube channel:

Converting XML to CSV with ITX

YouTube		remove preview
Converting XML to CSV with ITX In this practical demonstration, I will build a Typetree from a known CSV output requirement and an Input XML Schema for the same content: Use CSV example to import Typetree Modify that Typetree to suit requirements Mock-up an input XML Schema Create a map that reads the CSV and writes our XML Create a map that reads the XML and write to the CSV View this on YouTube >

#ITX  #XML  #CSV  #Transformation  #Importer  #Typetree

------------------------------
Paul Brett
 
IBM Sterling Transformation Extender (ITX) Client Support
------------------------------

Hi Paul, just a question, I was wondering what your view should be regarding using "infix" for the record delimiter and whether "postfix" would be a better choice? In order to ensure a file's integrity, and the "record / line" within it, I would inclined to assume that every valid line, should, by standard, be terminated by a <NL> sequence, otherwise if the file is corrupted, or truncated, it may still be processed without raising an error? However with "postfix", there is more likelihood that a corrupted, or truncated file would be detected, with "Input invalid" or "Input valid but unknown data found"? I'm just asking if this is useful to include in data integrity standards when processing CSV files? Or is there no potential gain either way?

------------------------------
Chris Clews
------------------------------

Infix is a personal choice.  I just prefer it, but then I'm not working with live data, and my input is always correct (in theory at least!).

Greg in our team favours the postfix style.

At the end of the day, each has it's own merits, and ultimately needs to reflect the data being sent.

------------------------------
Paul Brett
 
IBM Sterling Transformation Extender (ITX) Client Support
------------------------------

Original Message:
Sent: Thu May 08, 2025 08:19 AM
From: Chris Clews
Subject: Video - Converting XML to CSV with ITX

Hi Paul, just a question, I was wondering what your view should be regarding using "infix" for the record delimiter and whether "postfix" would be a better choice? In order to ensure a file's integrity, and the "record / line" within it, I would inclined to assume that every valid line, should, by standard, be terminated by a <NL> sequence, otherwise if the file is corrupted, or truncated, it may still be processed without raising an error? However with "postfix", there is more likelihood that a corrupted, or truncated file would be detected, with "Input invalid" or "Input valid but unknown data found"? I'm just asking if this is useful to include in data integrity standards when processing CSV files? Or is there no potential gain either way?

------------------------------
Chris Clews
------------------------------