Overview
Scans any pasted text — logs, markdown, email exports, meeting notes — and pulls out every URL it can find, either as a raw list in order of appearance or deduplicated down to unique hosts.
How It Works
Paste the text and the Matches counter shows the hit count immediately. All URLs returns every match in the order it appeared; Unique hosts only collapses the list to distinct hostnames sorted alphabetically, with mailto: addresses reduced to their bare addresses. Copy the list from the output panel.
Step-by-Step Usage Guide
- Paste logs, markdown, or any document into the text area.
- Choose All URLs for a verbatim list, or Unique hosts only for a sorted inventory of destinations.
- Check the Matches counter against expectations — a surprisingly low count usually means scheme-less links.
- Copy the extracted list.
Technical Specifications & Standards
Matching is a single case-insensitive pattern over http://, https://, and ftp:// URLs plus mailto: addresses. The pattern stops at whitespace and at the characters < > " ' ) ] — which is what keeps a URL wrapped in parentheses or angle brackets from dragging trailing punctuation along with it. Host mode parses each match with the URL API, keeps the port when present (localhost:3000 stays localhost:3000), and silently skips anything malformed.
Targeted Use Cases
- Auditing a markdown document or README for every external link before publishing.
- Pulling endpoints out of server logs or crawl output to see which services a system talks to.
- Collecting the addresses an email thread or newsletter export links to.
- Building a quick inventory of the domains referenced across a set of notes.
Notes & Gotchas
- Scheme-less links are not matched: www.example.com and bare addresses like contact@example.com are skipped by design — the sample text demonstrates both. Prepend a scheme, or use host mode, which reduces mailto: entries to plain addresses.
- All URLs keeps duplicates in input order. For an inventory, switch to host mode instead of eyeballing the long list.
- A trailing closing parenthesis is not captured, so a URL that legitimately ends in ) — some Wikipedia links — loses that character. Re-add it by hand if needed.
- Query strings are kept as-is; there is no strip-tracking-parameters option. Trim them in the copied output when you need canonical links.
Frequently Asked Questions
Does it deduplicate full URLs?
Only hosts mode deduplicates, and it does so per hostname. The All URLs list preserves every match in order, duplicates included.
Are mailto: links handled?
Yes. They appear in the All URLs list, and in host mode each one becomes the bare email address, sorted in with the hostnames.
Why did www.example.com not match?
It has no scheme. The pattern requires http://, https://, or ftp:// (or mailto:), which prevents false positives on plain words that happen to contain dots.
Is there a size limit?
No hard cap — matching runs in the browser on whatever you paste, bounded only by memory.