Regular expressions from zero¶

Clean 03 - Paper_Boats (Live 2024) into Paper Boats, while learning why each piece matches. This lesson takes about 20 minutes. Use a practice copy and the live Tag → Tag preview; save only after the whole result makes sense.

A copied audio file with TITLE = 03 - Paper_Boats (Live 2024).
Pending edits saved or discarded.
Familiarity with placeholders and function arguments from Format strings from zero.

1. Give the pattern one small job¶

Open Convert → Tag → Tag… (Alt+5; Alt is ⌥ Option on a Mac), set Field to COMMENT, and enter:

$regexp(%title%,'_',' ')

The preview should show 03 - Paper Boats (Live 2024).

Read the three arguments separately:

Argument	Job
`%title%`	Supply the text to inspect
`'_'`	Match the underscore character
`' '`	Replace every match with one space

A regular expression is a pattern describing the text to match. $regexp searches the input and replaces matches. An underscore has no special pattern meaning, so this first example behaves like literal replacement. Keep Field as COMMENT for now so the source title stays available while you experiment.

2. Recognize a changing number¶

Replace the expression with:

$regexp(%title%,'^\d+',)

The result is - Paper_Boats (Live 2024), including the leading space. The final argument is empty because the goal is deletion. Do not substitute '': that means one literal apostrophe in a format string.

Pattern part	Meaning	What it matches here
`^`	Start of the input	Position before `0`
`\d`	A digit	`0` or `3`
`+`	One or more of the preceding item	Both digits, `03`

The + lets the same pattern handle 3, 03 or 103. The ^ protects numbers elsewhere, including the year at the end.

3. Include the separator in the match¶

Deleting the number left its separator behind. Replace the pattern with:

$regexp(%title%,'^\d+\s*-\s*',)

The result is Paper_Boats (Live 2024).

\s* means zero or more whitespace characters. The first one consumes spaces between the number and hyphen; the second consumes spaces after the hyphen. - is literal here. Requiring the hyphen helps distinguish a track prefix from a title beginning with a meaningful number.

Input	Pattern	Replacement	Result
`03 - Paper_Boats (Live 2024)`	`^\d+\s-\s`	Leave empty	`Paper_Boats (Live 2024)`
`3-Paper Boats`	`^\d+\s-\s`	Leave empty	`Paper Boats`
`2024 Nights`	`^\d+\s-\s`	Leave empty	`2024 Nights`

In the last row, the required hyphen is absent. No match leaves the input unchanged; it is not a regex error.

Try it: require at least one space

Which quantifier would require whitespace instead of allowing none?

Answer

\s+ requires one or more whitespace characters. \s* permits zero.

4. Capture text to reuse it¶

Instead of deleting a prefix, capture the pieces:

$regexp(%title%,'^(\d+)\s*-\s*(.*)$','$2')

The result is again Paper_Boats (Live 2024). The round brackets create numbered capture groups:

Group	Captured text
`$1`	`03`
`$2`	`Paper_Boats (Live 2024)`

.* captures zero or more of any ordinary character. The final $ marks the end, so the pattern describes the whole title. $2 in the replacement copies the second captured group. The quotes around it protect the dollar sign from the outer format-string parser.

For comparison, replacing with '$2 — track $1' produces Paper_Boats (Live 2024) — track 03. Capture references rearrange matched text; they do not read tag fields.

5. Match literal brackets and a year¶

The version suffix is another independent job:

$regexp(%title%,'\s+\(Live \d{4}\)$',)

The result is 03 - Paper_Boats.

Pattern part	Why it is present
`\s+`	Remove the space before the suffix too
`$` and `$`	Match actual round brackets rather than create a group
`Live`	Match that word and its following space
`\d{4}`	Match exactly four digits
`$`	Require the suffix at the end

OpenTag ignores case by default in $regexp, so live also matches. Add a fourth argument 0 when the case must be significant: $regexp(%title%,'\s+$Live \d{4}$$',,0).

This pattern removes a very specific suffix. It will not remove (Live) without a year or (Studio 2024). Add new rules only after deciding those values should also change.

6. Combine the three verified steps¶

Use the prefix and suffix cleanup as nested functions, then change underscores into spaces:

$replace($regexp($regexp(%title%,'^\d+\s*-\s*',),'\s+\(Live \d{4}\)$',),_,' ')

Read from the inside outward:

Step	Intermediate text
Read `%title%`	`03 - Paper_Boats (Live 2024)`
Remove track prefix	`Paper_Boats (Live 2024)`
Remove the version suffix	`Paper_Boats`
Replace underscores	`Paper Boats`

The preview should now show Paper Boats. Keeping the jobs separate made the combined expression explainable; it also makes it easier to find which step needs changing when a new naming convention appears.

7. Look around without consuming text¶

For one more pattern technique, try this standalone expression:

$regexp(%title%,'\d{4}(?=\)$)',YEAR)

It produces 03 - Paper_Boats (Live YEAR). (?=\)$) is a lookahead: the digits must be followed by a closing bracket at the end, but that bracket is checked rather than included in the match. It therefore survives the replacement without being copied back.

Lookarounds are useful when the surrounding context decides whether a match is valid but should remain unchanged. Use an ordinary capture group when that communicates the task more clearly.

8. Apply the reviewed cleanup¶

Return to the combined expression in step 6. Change Field from COMMENT to TITLE, wait for the preview and check the exact output. Choose Apply to … files to save Paper Boats into the title. Inspect the result and use Edit → Undo to reverse the recorded practice change in the current desktop session.

Before applying the expression to a collection, test copies that include missing prefixes, unusual punctuation and suffixes you want to retain. Successful matching on this lesson's one input is not proof that every title in a library follows the same convention.

What you learned¶

Match fixed characters first, then describe the parts that vary.
Use anchors to protect unrelated numbers elsewhere in the title.
Include unwanted separators in a deletion pattern.
Capture groups save text for $1, $2 and later references.
Escape literal regex punctuation and quote patterns inside $regexp.
Lookarounds check context while leaving that context outside the match.
Verify each transformation before nesting or saving it.

Pattern part	Why it is present
`\s+`	Remove the space before the suffix too
`\(` and `\)`	Match actual round brackets rather than create a group
`Live`	Match that word and its following space
`\d{4}`	Match exactly four digits
`$`	Require the suffix at the end