Find and Replace

In plain text mode, whatever you type in the find box is escaped before being used, so characters that mean something to a regular expression — dots, brackets, question marks — are matched literally. Searching for 3.50 finds exactly that, not “3” followed by any character followed by “50”.

Whole words only wraps your search in word boundaries. Without it, replacing cat with dog also turns category into dogegory — the classic way a bulk replace quietly ruins a document.

Regular expression mode passes your pattern straight through. That unlocks things plain search cannot express — \d{4} for any four digits, \s+$ for trailing whitespace — and lets the replacement refer back to captured groups with $1 and $2. An invalid pattern is reported rather than silently doing nothing.

The match count updates as you type, before you commit to anything. That number is the safety check: if you expected to change three occurrences and it says 340, something in your pattern is wrong.

You have a 40-page policy document where the company name changed from Acme Ltd to Acme Group.

Paste it, find Acme Ltd, replace with Acme Group. The counter reads 27 matches. Copy the result — done in seconds, and nothing was uploaded, which for an internal policy document is the point.

Now a case where the naive approach fails. You want to change form to document. Type it and the counter reads 143 — far too many, because it is also matching format, information, performance and transform. Switch on whole words only and the count drops to 19, which is the real number.

A regex example. Your text has dates written as 2026-07-27 and you need 27/07/2026. Switch to regular expression mode, find (\d{4})-(\d{2})-(\d{2}), replace with $3/$2/$1. Every date flips, and the three captured groups are reordered by the replacement.

If your goal is tidying up stray spaces rather than changing words, the whitespace cleaner handles that in one step without needing a pattern.