Email Extractor

What the address pattern accepts

The scan is one regular expression, not a mailbox lookup. The local part may contain letters, digits, and the characters . ! # $ % & ' * + / = ? ^ _ ` { | } ~ -. The host is labels of letters, digits, and inner hyphens, and there must be at least one dot. user+tag@example.co.uk matches, and a trailing period after the last label is left outside the match. A comma, a closing angle bracket, or the word mailto: in front of an address is not part of the match, so the address itself still counts.

admin@localhost does not match, because there is no dot. a@b does not match either. ops@192.168.0.1 does match: the labels are numeric and the dots satisfy the pattern. That does not mean the address is deliverable. The page never looks up DNS, never opens a link, and never checks that the local part is a real person. Obfuscated wording such as a name followed by the word at and a host is ignored, because those characters are not an @.

Unique only and Lowercase results

Both boxes start checked. Lowercase results maps each match with toLowerCase before anything else. Unique only then walks that list and keeps a match only when its lowercased form has not been kept yet. The comparison key is lowercased even when Lowercase results is cleared, so Hello@Toolbench.online followed by hello@toolbench.online keeps only the first spelling. Clear Unique only and the second spelling remains. Clear both boxes and you get every match in the original case, in the order they occurred.

If Lowercase results stays on and Unique only is off, both copies are lowercased, so the list can show the same lowercase line twice. The count in the status is the number of lines printed, not the number of raw matches before unique filtering. Load sample does not extract by itself. Extract emails is the click that fills the pane. Changing a checkbox after a run leaves the old lines until you extract again.

When a line list of addresses is the goal

Use this scan when a thread, a footer, or a dump already contains real @ addresses and you want those strings on their own lines. It is not a line cleaner. Repeated non-address lines, wrapped sentences, and blank rows are simply ignored. If the next step is to drop duplicate lines of some other kind, paste the result into the text deduplicator. If the lines arrived wrapped, line operations joins breaks. Neither of those pages understands an @ pattern.

A literal find-and-replace is the find and replace tool. Sorting the extracted lines is the alphabetical line sorter. This page does not sort. First-seen order is the order of the output, which is why support@example.com follows hello@toolbench.online in the sample even though s comes before h.

Duplicates, punctuation, and hosts without a dot

Unique only is case-insensitive and order-preserving. It does not merge addresses that differ by a plus tag. ada@example.com and ada+list@example.com are two lines. It does not strip dots inside the local part. A match that ends at a dot, because the next character cannot extend the host, keeps the host without that dot. Punctuation glued to the front, such as an opening angle bracket, is outside the pattern and does not block the find.

An empty or all-whitespace box throws before the scan and the status is Paste text that contains email addresses. The pane is cleared. A box with text but no match succeeds: the status is No email addresses found. and the pane is empty, so Copy output has nothing to copy. That empty success is different from the error on a blank box. The page does not save the paste. Download output writes email-extractor.txt. There is no second column for the surrounding sentence, and no button that copies only the first address.

Rules and limitations

Matches need an @ and a dotted host. Unique only starts checked and drops later spellings of the same lowercased address. Lowercase results starts checked. First-seen order is kept.

A blank paste is an error and clears the list. Text with no match is a success with an empty list and the sentence No email addresses found. Download output saves email-extractor.txt. The page does not look up DNS.

Common problems

If a second spelling survived, Unique only was off, or the local part actually differed, including a plus tag. The uniqueness key does not ignore tags.

If admin@localhost is missing, the host has no dot, so the pattern skipped it. A numeric dotted host such as ops@192.168.0.1 is kept even though it is not a domain name.

Frequently asked questions

About this tool

Maintained by Toolbench. For large or deeply nested files, split the input into smaller batches and verify representative results before using them in an import or application. Report reproducible problems through Contact.