Links hide in text the same way email addresses do. A page of HTML source with hundreds of anchors. A chat export full of shared articles. A document where references are scattered through paragraphs. A server log where the URL is one field among many.
This extractor finds every http and https address in whatever you paste and gives you back a clean list.
How it decides what a link is
The pattern looks for http:// or https:// followed by everything up to whitespace or a character that cannot appear in a URL - a quote mark, an angle bracket, a closing parenthesis.
Trailing punctuation is trimmed, because a URL at the end of a sentence collects a full stop that is not part of the address. This is the single most common way extracted links end up broken.
One deliberate limitation: protocol-relative links written as //example.com/page and root-relative links written as /about are not matched. They are not complete addresses, and without knowing the page they came from there is no way to resolve them. If your HTML uses relative paths, this will find only the absolute ones.
Filtering
By protocol narrows the results to HTTPS or HTTP. The HTTP filter is the more useful one and has a specific job: finding insecure links on a page that is otherwise secure. Mixed content is blocked by browsers and is a common cause of a padlock disappearing after a migration to HTTPS.
By domain keeps only links whose hostname contains the text you enter. Type example.com and internal links are separated from external ones. This is the fastest way to build an internal link map from a page, or to check which third-party domains a page is reaching out to.
Unique and sort behave as elsewhere. Sorting groups by protocol first and then alphabetically, which naturally clusters links by domain and makes the shape of a page's outbound linking obvious.
Practical uses
The common ones are auditing. Extract the links from a page and run them through a link checker to find broken ones. Compare two sitemaps by extracting both and de-duplicating to find what changed. Pull the URLs from a chat export to build a reading list. Check a competitor's page for which domains it links out to.
For an internal audit, the domain filter does the work: extract everything, filter to your own domain, and you have the page's internal link structure without the noise of every social share button and external reference.
Nothing is uploaded. Server logs and internal page source frequently contain paths, parameters and identifiers that should not be pasted into a third-party service, which is the reason this runs locally.