Notepad++ Capitalize Every First Letter of Every Word

Background

According to Notepad++ specification (see Substitutions section), there are three operators that can be useful when turning substrings uppercase:

\u
Causes next character to output in uppercase
\U
Causes next characters to be output in uppercase, until a \E is found.
\E
Puts an end to forced case mode initiated by \L or \U.

Thus, you can either match a substring and turn its first character uppercase with \u to capitalize it, or match a character and use \U/\E.

Note that Unicode characters won't be turned uppercase, only ASCII letters are affected.


BOW (Beginning of Word) Bug in Notepad++

Note that currently (in Notepad++ v.6.8.8) the beginning of word does not work for some reason. A common solution that works with most engines (use it in Sublime Text and it will match) does not work:

\b(\w)

This regex matches all word characters irrespective of their position in the string.

I logged a bug Word boundary issue with a generic subpattern next to it #1404.


Solution #1 (for the current Notepad++ v.6.8.8)

The first solution can be using the \w+ and replace with \u$0 (no need using any capturing groups). Though this does not mean we only match the characters at the beginning of a word, the pattern will just match chunks of word characters ([a-zA-Z0-9_] + all Unicode letters/digits) and will turn the first character uppercase.


Solution #2 (for the current Notepad++ v.6.8.8)

The second solution can be implemented with special boundaries defined with lookbehinds:

(?:(?<=^)|(?<=\W))\w

And replace with \U$0\E.

The regex (?:(?<=^)|(?<=\W))\w matches an alphanumeric only at the beginning of a line ((?<=^)) or after a non-word character ((?<=\W)).

The replacement - \U$0\E - contains a \U flag that starts turning letters uppercase and \E is a flag that tells Notepad++ to stop converting case.


Edge case

In case you have hyphenated words, like well-known, and you only want the first part to be capitalized, you can use [\w-]+ with \u$0 replacement. It will also keep strings like -v or --help intact.


A simpler regex that worked for me:

Find: (\w+)

Replace: \u$0