Detect URL in text without hyperlink and replace normal URL with hyperlink?

Question

I have spent considerable amount of time searching for the solution or trying one, But I did not found one. So my usecase is:

I have a text which can have simple url(with or without http/s) or it can also have hyperlinked url.

What regex should do

It should leave hyperlink url as it is and convert the non hyperlinked url to a hyperlinked URL.

I am learning regex from www.codeburst.com and trying regex at <a href="https://regexr.com">Regexr</a>.

I am learning regex from <a href="www.codeburst.com">www.codeburst.com</a> and trying regex at <a href="https://regexr.com">Regexr</a>.

this regex, but it it not working as expected.

/((?!href).((https?:\/\/)||(www\.)|(mailto:)).+)/gi

I don't see a difference between your source text and your 'Expected Solution' In what way does your regex not do what you want? — MikeB, May 02 '19 at 09:52
Made changes to the question, now html part is showing correctly. — Sachin Jagtap, May 02 '19 at 09:56
Possible duplicate of [What is a good regular expression to match a URL?](https://stackoverflow.com/questions/3809401/what-is-a-good-regular-expression-to-match-a-url) — Plochie, May 02 '19 at 09:58
No. Both questions are different. There the questioner wants regex which only pass if url start with http or https. Mine url can start with http, https, www and mailto also. — Sachin Jagtap, May 02 '19 at 10:00

Jan Turoň · Answer 1 · 2019-05-02T12:02:33.207

1

You probably need a negative lookbehind (?<!href=") which was added to ECMAScript recently, see this answer
be careful with double || which renders tokend behind this useless (hungry match)
also be careful with .+ which matches everything after (including newline with /s regex option)

I would start with

(?<!href=")(((https?:\/\/)|(www\.)|(mailto:))\S+)

edited May 02 '19 at 12:02

answered May 02 '19 at 11:55

Jan Turoň

score 0 · Accepted Answer · answered May 02 '19 at 11:30

0

(https?:\/\/)?(www\.)?[-a-zA-Z0-9@:%._\+~#=]{2,256}\.[a-z]{2,6}\b([-a-zA-Z0-9@:%_\+.~#?&//=]*)

The breakout of regex is as follows,

I have checked above regex with following test cases:

answered May 02 '19 at 11:30

Plochie

Your solution works correctly for urls without hyperlink, but if text is 'with hyperlink url ABC', then it gets converted to 'with hyperlink url https://www.example.com">ABC'. For urls inside href, it should keep them as it is. – Sachin Jagtap May 02 '19 at 11:34
1

You can search all the strings matching with above url, and then find the strings not starting with `href` programmatically. – Plochie May 02 '19 at 12:30