# Roman Numerals case conversion
**URL:** https://community.mp3tag.de/t/roman-numerals-case-conversion/5520
**Category:** Support
**Created:** [October 1, 2007, 8:47am UTC](https://community.mp3tag.de/t/roman-numerals-case-conversion/5520 "2007-10-01T08:47:07Z")
**Posts on this page:** 7
**Page:** 1
### Author:  [@sminc](https://community.mp3tag.de/u/sminc)
#### Post date: [October 1, 2007, 8:47am UTC](https://community.mp3tag.de/t/roman-numerals-case-conversion/5520/1 "2007-10-01T08:47:07Z")
I'm using the regex example to convert roman numerals to upper case but I'm having a problem with it.
For some reason it is deleting commas out of the tag / filename / directory which makes no sense as the only replace command is the $upper
The suggested regex is
> [@](#):
>
> Upper case for Roman numbers
>
> Regular expression:
> (^|\W)(?:M{0,3})(?:D?C{0,3}|C[DM])(?:L?X{0,3}|X[LC])(?:V?I{0,3}|I[VX])(?=(.\s|\s|)|$))
> Replace matches with:
> $upper($0)
> case-sensitive comparison
Is there something in here where the comma is being selected but the $upper fails so it is being replaced with nothing ?
---
### Author:  [@dano](https://community.mp3tag.de/u/dano)
#### Post date: [October 1, 2007, 12:01pm UTC](https://community.mp3tag.de/t/roman-numerals-case-conversion/5520/2 "2007-10-01T12:01:40Z")
The problem is when $upper gets a parameter with a comma it takes it as seperator and not as string.
If you replace at the begining (^|\W) by \b it should avoid this bug.
---
### Author:  [@egd](https://community.mp3tag.de/u/egd)
#### Post date: [May 9, 2008, 7:25pm UTC](https://community.mp3tag.de/t/roman-numerals-case-conversion/5520/3 "2008-05-09T19:25:57Z")
Dano, am I correct in assuming that should be replace (^|\W) with \b ?
---
### Author:  [@DetlevD](https://community.mp3tag.de/u/DetlevD)
#### Post date: [May 10, 2008, 4:26am UTC](https://community.mp3tag.de/t/roman-numerals-case-conversion/5520/4 "2008-05-10T04:26:57Z")
If you write a regular expression, which contains a syntactical correct comma in the search string, within Mp3tag $regexp() function you have to escape the comma with apostrophes, because the comma is the delimiter character in Mp3tag scripting language that separates the parameters in a parameterlist.
DD.20080510.0827.CEST
---
### Author:  [@dano](https://community.mp3tag.de/u/dano)
#### Post date: [May 10, 2008, 6:26am UTC](https://community.mp3tag.de/t/roman-numerals-case-conversion/5520/5 "2008-05-10T06:26:04Z")
egd I've updated the expression in the thread.
---
### Author:  [@chrisjj](https://community.mp3tag.de/u/chrisjj)
#### Post date: [May 13, 2008, 5:27pm UTC](https://community.mp3tag.de/t/roman-numerals-case-conversion/5520/6 "2008-05-13T17:27:02Z")
> [@](#):
>
> when $upper gets a parameter with a comma it takes it as seperator and not as string.
Yuk. Especially as $upper has nothing to separate, being a single-parameter function.
Thanks for that info, Dano.
> [@](#):
>
> If you replace at the begining (^|\W) by \b it should avoid this bug.
What's \b ? Not mentioned in the Help, AFAICS.
---
### Author:  [@system](https://community.mp3tag.de/u/system)
#### Post date: [February 9, 2026, 12:20pm UTC](https://community.mp3tag.de/t/roman-numerals-case-conversion/5520/7 "2026-02-09T12:20:31Z")