Hi,
Using the Regex tool in the Parse category, if you set the Output Method to "Parse", the following RegEx should do what you want (I have it set up so that it outputs the 0 when it appears, but you can apply replacement logic afterwards to handle this).
_(0|\d{9})_(\d{9})
I will edit this shortly to break down the RegEx and explain it.
EDIT: RegEx breakdown:
_(0|\d{9})
This is the first underscore, followed by either a 0 or 9 digits.
The parentheses create the first marked group that will be parsed out, so you can move these around to add the underscore if needed.
The "|" character signifies an "OR" function.
The \d{9} is 9 digit characters.
The second marked group is very similar to the first, but we have no 0's in the example so I did not include the 0| logic. You could add this if it exists in your test case.
Second EDIT:
If you want to avoid using a formula to replace the 0's that appear later, the following RegEx looks to be a little better.
_(?:0|(\d{9}))_(\d{9})The main difference here is that we have isolated the first marked group to \d{9}, and used an Unmarked group for the 0 OR check. This way, 0's will find a match, but will not appear in the final output, giving a blank value.
One thing to note is that line 8 which was not highlighted in your example also has matched output.
