Regular Expressions

[Florian]

This thread collects some useful regular expressions.

If you created a regular expression which solves a common task, please post it here and give a short description what it is supposed to do.

These regular expressions can be used with the action type Replace with regular expression. Please look at FAQ: How do I create a new action? to learn more about actions in Mp3tag.

-> If you have problems with a regular expression, please open a separate topic.

[Florian]

Trim leading/trailing spaces

See FAQ: Trim leading/trailing spaces.

[areve]

Regular expression: ^\s*[0-9]+\s*-\s*
Replace with:

This one will remove the track-number (if followed by a dash) and white space at the beginning of a string
(for instance 01 - Come Together will become Come Together)

Regular expression: ^\s*[0-9]+\s+
Replace with:

This one will remove the track-number if followed by a white space
(for instance 01 Come Together will become Come Together)

[areve]

Here is another one.
I don't think it will be useful as such to a lot of people, but with REGEXP, examples are never too many

My files were tagged as such:

name of the artist.2003AL-name of the album

(the 2003AL meaning: released in 2003, Album, Live, but it could be 1977A, or 1995IR)
Weird, I know
I wanted to change that to a simpler

2003-name of the album

Here is what I did:
Regular expression: ..([0-9]).*-(.+)
Replace with: $1-$2

[phoenixdarkdirk]

This regular expression fixes tracknumbers from iTunes in the [track]/[numtracks] format (like 3/12).

Regex:

^(\d+)/\d+

Replace with:

$1

This can be applied to every file and it will only correct ones that have the / in them. Good luck!

According to RevRagnarok's suggestions below, I've made the expression less greedy.

[areve]

Another one, that switches first and last names (for instance, Jacques Brel will become Brel, Jacques):
Regular Expression: ^(.+)\s(.+)$
Replace with: $2, $1

[mll]

Used to use this basic regexp :

Applied on TRACK, find ^(\d)$ and replace by 0$1

It adds a leading 0 to the track number. I was thinking about enhancing it to apply it only to 1-digit numbers... when I discovered Ctrl+K :

Cheers,

MLL

According to RevRagnarok's suggestions below, I've made the expression less greedy

[ThurstonX]

If you have a filename:

• 01 Trackname

and want it to be

• 01 - Trackname

create an Action with the following properties:

• Field: _FILENAME
• Regular Expression: ([0-9]+)\s
• Replace matches with: $1 - $2

Edit: Topic merged by moderator.

[RevRagnarok]

I've noticed a lot of these RegExs 'go overboard' in their matching. A key to good expressions are limiting (1) the false matches and (2) the how long the engine needs to analyze the string.

For example, my two favorite expressions I have posted here are both above on this page. However, they have very greedy matchers that can easily result in lost data.

phoenixdarkdirk's - why NOT limit the track 'numbers' to digits with \d+ ?
mll's - really went overboard when a simple ^(\d)$ would've done it.

I highly recommend this book if you are serious about using REs. Of course, I recommend trying to find it at a technical library, because if you look it over a little you may realize you were just kidding and save yourself the money.

Of course, when we are using them on a handful of MP3 files, it's no big deal. It's when you are handling megabytes of text files that it really matters.

Edit 5 Apr 2013 - After over eight years, fixed the URL to my blog.

[Florian]

RevRagnarok,

I've noticed a lot of these RegExs 'go overboard' in their matching. A key to good expressions are limiting (1) the false matches and (2) the how long the engine needs to analyze the string.

Thanks a lot for your suggestions. I've changed the expressions you've mentioned to be less greedy.

Best regards,
~ Florian

[nickless]

This RegEx will convert abbreviations composed of single chars and points between (and/or behind it) to uppercase.
Single chars without points around remains lowercase, except if a "-"char and a space is before it. (Example: "Songname - A text")

RegEx: ( |\.|^)(\w)(?= |\.|$)
Replace with: $1$upper($2)and
RegEx: ([^-])( \u )
Replace with: $1$lower($2)

Example: a.b a Reg.eX in p.o.d p. diidi e.t. - a bad a.i

==>
A.B a Reg.eX in P.O.D P. diidi E.T. - A bad A.I

Both regular expressions should be executed one after another

Use

RegEx: ( \u )
Replace with: $lower($1)instead of second RegEx to have chars after " - " lowercase too. (Example: "Songname - a text")

Regards
nickless

Edit: removed some unnecessary characters from RegEx

[ThurstonX]

Title Tag Conversion

This 2-step action does:

1. set Title tag to Filename
2. convert titles using, e.g.:
   01 - Track Name
   to
3. Track Name

First step:
Format Value
Field: TITLE
Formatstring: %_FILENAME%

Second step:
Set a new Replace with Regular Expression action
Field: TITLE
Regular expression:
^([0-9]+)\s*-\s*
Replace matches with:
$1. $2

I use the first format for all file names, but my iRiver displays the title tag, which looks better (and saves one character) using the latter. For me it's a 2-step process: Filename-to-Tag (%TITLE% only) and then convert Title tag. Assumes your filename is the way you want it, of course.

[dano]

Upper case for Roman numbers

Regular expression:
\b(?: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

[Snykch]

Convert artist names in the form 'Artist, The'

Regular expression: (.*),\sthe$
Replace with: The $1

[Michaelm]

And the other way around

Move "The " to the end

Regular expression: ^The (.+)
Replace with: $1, The

So "The Cure" becomes "Cure, The"
And "The a whole lot of words" becomes "a whole lot of words, The"

[Blues_Fan]

It's taken me far too long to figure it out but I've finally got it working.

To change Horne, Lena to Lena Horne or
Hawkins, Coleman With Manny Albam & His Orchestra to Coleman Hawkins With Manny Albam & His Orchestra
Field: ARTIST
Regular Expression: ^([\w]+),\s([\w]+)

Replace Matches with: $2 $1

I hope I didn't miss the answer in here somewhere. If there's an easier or better way please feel free to help.

[neilpa]

This regex can remove/replace all but the last dot from a string. It's useful for people like me who prefer to limit filename chars to a-z, 0-9, '_', and '-'. We need to save that last dot for file extensions.

Regex: .

Assumes file extensions are 2-4 chars long (not including the dot). If you are curious how this works google regex lookaround.

[nmiller_co]

I recently converted all my wma files to mp3 only to realize that all the tags were lost. My music is in the form:

...\artist\album\track title.mp3

MP3TAG came to my rescue! Not sure whether there was an easier way to do this, but I got started by reading this. Sorry if these are lame ways of doing it, but it's my 1st experience with regexp.

format artist:
$regexp(%_folderpath%,.+\\(.+)\\(.+)\\,$1)

format album:
$regexp(%_folderpath%,.+\\(.+)\\(.+)\\,$2)

Next 2 I got from ThurstonX in this post.

format track:
$regexp(%_filename%,(\d*)\s?(.+),$1)

format title:
$regexp(%_filename%,(\d*)\s?(.+),$2)

This program saved me TONS of time...

[Harakiri]

Another member helped me creating one of the most useful REGEXP that i use so far:

From this:

Röyksopp - Beautiful DAY withOUT yoU (Rex The Dog Remix)

To this:

Röyksopp - Beautiful day without you (Rex The Dog Remix)

USING:

Action type: Replace with regular expression
Field: TITLE
Regular expression: ^(.*?)((|$)
Replace matches with: $caps3($1)$2
[ ] case-sensitive comparison

See the original thread here

[KCE]

Really simple but this will remove any parenthesis and its contents in any part of the string

Example:
Blah (blah1) (blah 2)
Result:
Blah

Regular expression: (.+?)

Replace matches with: