Generic Object Oriented Solution
For people like me that use frameworks like angular that don't allow manipulating DOM directly, I created a function that takes a string and returns an array of url/plainText objects that can be used to create any UI representation that you want.
URL regex
For URL matching I used (slightly adapted) h0mayun regex: /(?:(?:https?:\/\/)|(?:www\.))[^\s]+/g
My function also drops punctuation characters from the end of a URL like . and , that I believe more often will be actual punctuation than a legit URL ending (but it could be! This is not rigorous science as other answers explain well) For that I apply the following regex onto matched URLs /^(.+?)([.,?!'"]*)$/.
Typescript code
export function urlMatcherInText(inputString: string): UrlMatcherResult[] {
if (! inputString) return [];
const results: UrlMatcherResult[] = [];
function addText(text: string) {
if (! text) return;
const result = new UrlMatcherResult();
result.type = 'text';
result.value = text;
results.push(result);
}
function addUrl(url: string) {
if (! url) return;
const result = new UrlMatcherResult();
result.type = 'url';
result.value = url;
results.push(result);
}
const findUrlRegex = /(?:(?:https?:\/\/)|(?:www\.))[^\s]+/g;
const cleanUrlRegex = /^(.+?)([.,?!'"]*)$/;
let match: RegExpExecArray;
let indexOfStartOfString = 0;
do {
match = findUrlRegex.exec(inputString);
if (match) {
const text = inputString.substr(indexOfStartOfString, match.index - indexOfStartOfString);
addText(text);
var dirtyUrl = match[0];
var urlDirtyMatch = cleanUrlRegex.exec(dirtyUrl);
addUrl(urlDirtyMatch[1]);
addText(urlDirtyMatch[2]);
indexOfStartOfString = match.index + dirtyUrl.length;
}
}
while (match);
const remainingText = inputString.substr(indexOfStartOfString, inputString.length - indexOfStartOfString);
addText(remainingText);
return results;
}
export class UrlMatcherResult {
public type: 'url' | 'text'
public value: string
}