Extraction

Extraction reads the pages of every site in a cluster and pulls out the parts that match an expression. A cluster of sites becomes a table of phone numbers, social accounts, addresses, plugin names - whatever the pattern captures.

What it reads

Everything PublicWWW has indexed for those sites, internal pages included - not only the home page a search matched on. A contact detail that only appears on an "about" page is still found.

Presets and your own patterns

There are presets for the things people ask for most - email addresses, phone numbers, WhatsApp, Telegram, Viber and LINE contacts, links to Facebook, Instagram, X, LinkedIn (people and companies), YouTube, TikTok, Pinterest, Threads, Bluesky, Discord and GitHub profiles, the site's apps in Google Play and the App Store - and any regular expression can be used instead. The expression works the same way snipexp: does in a search: the capturing group is what ends up in the column, and an expression without one produces empty results.

|/wp-content/plugins/([\w\d\-\.\_]+)/|

Test on a small cluster first. The expression can be tried, adjusted and tried again with a cap on how many records to pull, so a wrong pattern costs almost nothing - see limits for why that matters.

Getting the result out

The extracted table downloads as a file and opens in a spreadsheet. Each row is a site and what was found on it; a site where nothing matched is still listed, so the gaps are visible rather than silently dropped.

Filtering searches with a cluster

A cluster works as a filter in the other direction too. Upload your own list of domains, urls or email addresses, and use it to reduce search results to the sites on that list - or, by subtracting, to everything except them.

Extracting phones and emails walks the whole thing through, from two searches to a spreadsheet. From code, extraction is part of the API, with the same presets by name.