Filter a certain amount of letters out of a code

Daniel Elpunkt

So basically I am having a long list of codes with 12 characters, including A-Z and 0-9.
I want to delete every code which has more than 3 letters.
for exampe: C9RUEOCXK60E
has more than 3 letters - should be filtered out.
C95678C3210E has 3 letters - I want to keep this one.

How am i gonna do this?

gerdb42

This could be done with a RegEx: \b([[:alpha:]]+?[[:digit:]]*){4,}\b to search for and an empty String for replacement. If you want to delete the entire line, replace the second \b with \R.

What it does: after a word boundary (\b) look for the shortest possible sequence of alpha-characters ([[:alpha:]]+?), optionally followed by a sequence of digits ([[:digit:]]*) with at least four repetitions ({4,}) up to the next word boundary.

guy038

Hi, @daniel-elpunkt,

@gerdb42’s regex works fine. Just an other formulation, based on the fact that your codes do not contain lower-case letters :

SEARCH \b(\d*\u){4,}\d*\b

REPLACE EMPTY

And, as @gerdb42 said, if you have a single code, only, of 12 characters long, per line, just replace the final \b by the syntax \R, which represents any End of Line characters, whatever the file type ( Windows, Unix or Mac )

SEARCH \b(\d*\u){4,}\d*\R

REPLACE EMPTY

After replacement, only codes, containing 0, 1, 2 or 3 capital letters, are kept !

Best Regards,

guy038