Got it, I see the problem with your current regex! The /{{([^}]+)}}/g works great for standard double-brace cases, but when you hit triple braces (like {{{Text}}}), it messes up because [^}]+ doesn't care if there's an extra { inside the match—it just grabs everything until the first }. Let's fix this with lookaround assertions to make sure we only match exactly double braces, no more, no less.

The Updated Regex #

/(?<!{){{((?:(?!}}).)+)}}(?!})/g

Let's Break This Down #

Here's why this fixes your issue:

• (?<!{): This is a negative lookbehind—it checks that the {{ isn't immediately preceded by another {. So it won't accidentally match the first two braces in a {{{ sequence.
• {{: Just matches the opening double braces, plain and simple.
• ((?:(?!}}).)+): Instead of using [^}]+ (which stops at any single }), this part matches any character except when it's the start of a closing }} pair. That means if you have a single } inside your content (like {{Hello}World}}), it will still match the whole thing instead of cutting off early.
• }}: Matches the closing double braces exactly.
• (?!}): A negative lookahead that makes sure the }} isn't followed by another }. So it won't match the last two braces in a }}} sequence.

Test It Out #

• For your example { {{Text}} }, this regex will perfectly target {{Text}}—the spaces around the braces don't trigger the lookaround checks, which only care about adjacent { or } characters.
• If you have something like {{{TripleStart}}} or {{TripleEnd}}}, it won't match anything at all—those extra braces get blocked by the lookbehind/lookahead.
• Even trickier cases like {{This has a single } inside}} work correctly, since the regex doesn't stop at the first }.

Why Your Original Regex Was Acting Up #

The [^}]+ in your original regex matches any character that isn't a }, so when it sees {{{Text}}}, it grabs the third { along with Text (since { isn't a }), then pairs that with the first two closing }. That's why you ended up with a match that included the extra opening brace and missed the last one—this new regex fixes that by enforcing strict double-brace boundaries.

内容的提问来源于stack exchange，提问作者jonnoMK