urlsemailsnumbershashtagsmentionsevery find is painted in place — click a chip to copy it
NETWORK PAYLOAD: 0 BYTES · PROCESSING: LOCAL
Extraction channels0 MATCHES
Total matches0ACROSS ACTIVE CHANNELS
Chars scanned0ON EVERY KEYSTROKE
Categories hit0/5WITH AT LEAST ONE FIND
Density0.0MATCHES PER 100 WORDS
✱ Text Tools · Group 05 of 09
Text Extraction.
Five pattern hunters that pull URLs, emails, numbers, hashtags and mentions out of any text mess — and paint every find directly in the editor as you type. Click a chip to copy it, or grab a whole channel at once. Nothing is uploaded, nothing is stored.
Catches every link that starts with http://, https:// or www. — query strings and fragments included — and strips trailing sentence punctuation so a link followed by a comma comes out clean. Bare domains are deliberately ignored: without a scheme they're too easy to confuse with file names.
Example
Read https://example.com/guide, then visit www.example.com/faq.
↓
https://example.com/guide www.example.com/faq 2 links — the trailing comma was stripped from the first.
Use it for reference listslink auditssource collectiondead-link prep
Pulls every valid mailbox out of the text: local part, @, domain and a real TLD. Dots, plus signs, hyphens and underscores in the local part are all understood; decoys like "bob (at) example" are ignored because they aren't addresses.
Example
Mail bob@example.com or press@team.example.co.uk — not bob (at) example.
↓
bob@example.com press@team.example.co.uk 2 mailboxes — the "(at)" decoy is ignored.
Use it for contact listsCRM cleanuplead collectionsupport rosters
Catches integers, decimals and negatives — and only real numbers. Digits glued to letters ("v2", "room4") are skipped, and digits that belong to a URL or email stay with their parent pattern. What you get is the pure numeric content of the text, in reading order.
Example
Pi is 3.14159, it was -7 degrees, and room v2 holds 100.
↓
3.14159 · -7 · 100 — the "2" in "v2" stays put.
Use it for data scrapingprice listsstatistics gatheringinvoice digits
Finds every #tag built from letters, digits and underscores — in any script — while steering clear of lookalikes. HTML entities like ' are never mistaken for tags, and a # inside a URL belongs to the URL, not to the hashtag channel.
Example
Follow #launchday and #build_in_public, ignore 'entities'.
↓
#launchday · #build_in_public — entities left alone.
Use it for campaign trackingsocial listeningtag researchthread roundups
Finds every @handle — letters, digits and underscores — and knows the difference between a mention and the @ inside an email address. An @ glued to a word on its left ("bob@") is part of a mailbox, so it's claimed by the email channel instead.
Example
cc @ana_ops, ping @dev_team, mail team@example.com
↓
@ana_ops · @dev_team — the @ in the email stays where it belongs.
Use it for collaborator liststhread summariesattributioncommunity audits
Extraction rules · urls = http(s):// and www. starts, trailing punctuation stripped · emails = local@domain.tld · numbers = integers, decimals, negatives — never inside words, urls or emails · hashtags & mentions = #/@ + letters, digits, underscores, never inside urls or entities · overlaps resolved once: each span belongs to its strongest pattern · click any chip to copy it
The privacy promise
Paste anything. We couldn't read it even if we wanted to.
This page has no backend and no database. Every match you just saw highlighted was found by JavaScript inside your own browser tab — your text never traveled a single centimeter of cable. Close the page and it's gone for good.
YOUR TEXT→SCANNED IN YOUR BROWSER→YOUR FINDS
✕ SERVER · DATABASE · LOGS · ANALYTICS
Databasenull
Backendnone
Text uploads0
Session logs0
Questions
Asked often, answered plainly.
( 02 )
Why wasn't "example.com" extracted as a URL?
Only links that start with http://, https:// or www. are matched. Bare domains are skipped on purpose — without a scheme they're ambiguous and easily confused with file names or sentence fragments. Add the protocol and they'll be caught instantly.
Do numbers inside emails or URLs get counted?
No. Overlaps are resolved once, and each span of text belongs to exactly one pattern: URLs beat emails, emails beat mentions and hashtags, and those beat plain numbers. A digit inside an address is part of the address, nothing more — which is exactly what the highlight colors show you.
Can I copy everything from one channel at once?
Yes. Every channel header has a "copy all" button that copies the full list, newline-separated — even when the visible chips are capped on very long texts. And clicking any single chip copies just that find.
Is the text I paste saved anywhere?
No. IrisTool has no backend and no database. All extraction runs inside your browser tab and disappears the moment you close or refresh the page. There is nothing to leak, because nothing is ever collected.
Can I use these tools offline?
Yes. Once the page has loaded, every extractor on it keeps working offline, because all the logic runs locally in your browser.