Regex with Exiftool – Removing characters from a group question

[philbond87]

The forum recently helped me with this but I have to refine it.

The Exiftool command contains a regex that forms a string from three regex groups.
At the end of the command I remove certain characters from the string BUT what I really need to do is remove those characters only from the third group.

Here is my command, as it's currently implemented:

Code Select Expand

exiftool -overwrite_original_in_place -r '-transmissionreference<${filename;m/(\d{6}).*_([A-Z]{2,})_?([iwOSEV]*\d{6})/;$_=$1.$2.$3;s/w//;s/E//;s/O//;s/S//}'

I had thought that I was removing "w", "E", "O", "S" from the third group but I was actually removing them from the entire string.
How can I remove them from the third group only?

Thanks!

[philbond87]

I suppose this is more of a pearl question then.

I have tried this, which seems to work for getting one of the needed characters removed in the last group, but I can't figure out the syntax to get more than one character removed:

Code Select Expand

exiftool -overwrite_original_in_place -r '-transmissionreference<${filename;m/(\d{6}).*_([A-Z]{2,})_?([iwOSEV]*\d{6})/;$_=$1.$2.($3=~ s/w//r)}'

[philbond87]

I think I've found the solution.
I think I can simply stack the replacement characters in the single substitution expression with logical ors.

Code Select Expand

;$_=$1.$2.($3=~ s/w|E|O//r)}'

[StarGeek]

Glad you found a solution.  I didn't go through your regex but one thing to take note about your final post is that only the first of those letters will get removed.  For example, given WoOeEw, only the O will get removed.  To remove all matching characters, add g so the substitution will be global.  Also the first letter will not get removed because it is a capital W and the case insensitive modifier (i) isn't there.

[philbond87]

Thanks.

There will only be one of each of the selected characters in the third regex group (which is the only group I'm concerned with removing those characters from).
If I did want to account for multiple occurrences in the third group where would I put the 'g' global to the substitution expression?

Also I am only concerned with removing a possible, single lowercase 'w' so that should be fine.

Thanks again.

[philbond87]

@StarGeek,

I spoke a bit too soon. I misunderstood what you were saying about the global modifier.
Yes, I do need the global modifier to get all of the occurrences of the characters.

[StarGeek]

The modifiers for a Perl regex go at the end
s/w|E|O//rg