Regex to find and replace emoji names within colons

Question

I'm trying to write a regex (for JavaScript's regex engine) that I can use to do a find and replace in text for emoji names within colons. Like in Slack or Discord when you type :smiley-face: and it replaces it when you submit the chat. I'm targeting text nodes only so I don't need to worry about other html inside the text.

Is it possible to write a regex that could match all of the following rules? (text highlighted with monospace blocks = regex positive matches)

:any-non-whitespace:
:text1:sample2:
:@(1@#$@SD: :s:
:nospace::inbetween: because there are 2 colons in the middle
:nospace:middle:nospace:

I'm starting with something like this but it's incomplete

/:(?!:)\S+:/gim

I'm trying to think of all the special cases that might possibly occur doing this. Maybe I'm overthinking it.

There's a lot of Twitch emotes involved so I can't use emoji unicode characters. The regex will find matches and replace with tags

I am not quite clear about your rules plz come up with some examples for each of your rules — Mustofa Rizwan, Apr 10 '18 at 04:43
Your second example shouldn't be allowed as a valid emoji format. That would make this much easier. — user3483203, Apr 10 '18 at 05:07
You may try [`:[^:\s]*(?:::[^:\s]*)*:`](https://regex101.com/r/tjHseH/1). Same as [`:(?:[^:\s]|::)*:`](https://regex101.com/r/tjHseH/2), but a bit more efficient. — Wiktor Stribiżew, Apr 10 '18 at 05:08
@chrisz Actually OP does not need any. Why did you modify the OP requirements? — Wiktor Stribiżew, Apr 10 '18 at 05:46
Fransisc0, I rolled back the changes done by chrisz to your original post. Please check my suggestion and let know if either works. — Wiktor Stribiżew, Apr 10 '18 at 06:30
@WiktorStribiżew my mistake, believed it was a formatting error. — user3483203, Apr 10 '18 at 13:02

score 6 · Accepted Answer · answered Apr 11 '18 at 20:25

I suggest using

:[^:\s]*(?:::[^:\s]*)*:

See the regex demo. It is the same pattern as :(?:[^:\s]|::)*:, but a bit more efficient because the (?:..|...)* part is unrolled.

Details

: - a colon
[^:\s]* - 0+ chars other than : and whitespace
(?: - start of a quantified non-capturing group:
- :: - double colon
- [^:\s]* - 0+ chars other than : and whitespace
)* - end of grouping, repeated 0 or more times (due to the * quantifier)
: - a colon.

Thm Lee · Answer 2 · 2018-04-10T09:08:29.853

1

Do you want something like this regex?

(:(?![\n])[()#$@-\w]+:)

Demo,,, in which you can additionally insert unallowed characters into the character class of the (?![\n]) and also additonally insert allowed characters into the character class [()#$@-\w]

edited Apr 10 '18 at 09:08

answered Apr 10 '18 at 09:03

Thm Lee

1,236
1
9
12

This splits the double colon `::`. – SamWhan Apr 10 '18 at 12:24
@ClasG For your reference, leave this link https://demos.emojione.com/ascii-smileys.html – Thm Lee Apr 10 '18 at 14:57

SamWhan · Answer 3 · 2018-04-10T12:26:39.017

0

My first thought was

:(::|[^:\n])+:

It matches a string, at least one character long, including surrounding colons, that consists of either

two colons (::), or
a character that isn't a colon, nor a line feed.

But that's basically what Wiktor had as a (slower) alternative (comments). But I'll leave it here anyway since it's working, as opposed to the other submitted answers ;)

See it here at regex101.

edited Apr 10 '18 at 12:26

answered Apr 10 '18 at 08:36

SamWhan

8,296
1
18
45

score 0 · Answer 4 · edited Mar 18 '19 at 14:24

0

Try this regx

/(^|\s)+:([^\s\n\r])+:|^:[^\s\n\r]+/g

edited Mar 18 '19 at 14:24

Vishal Pawar

4,324
4
28
54

answered Mar 18 '19 at 13:40

Satya Pendem

307
5
13

1

Answering questions is good but your answer would be much better if it explained what the regex does and how it works. – AdrianHHH Mar 18 '19 at 14:03

Regex to find and replace emoji names within colons

4 Answers4

Linked

Related