Reference

Search syntax

SearchWebCode matches your query against the raw HTML source of 131M+ homepages. Combine exact phrases with AND / OR / NOT logic, filter by TLD, or drop into full regular expressions. Every operator below also works in the search box, or build it visually.

Open the search box  →

Operators

Whitespace between terms means AND, every term must appear on the page.

SyntaxMeaning
"exact phrase"Match the string exactly, including spaces and punctuation, wherever it occurs in the source. Quotes are optional for a single word but required for phrases or anything with a space.
wordA bare single word (letters, digits, -, _; quoted or not) matches as a whole word, case-insensitively, never inside a longer one: tagmanager does not find googletagmanager, and 2380747294665477 does not find ca-pub-2380747294665477. For the substring reading switch to regex, /tagmanager/, or include the punctuation that pins it: ca-pub-2380747294665477.
a  b  ·  a AND bAND, a space between terms requires both on the same page; the uppercase keyword means the same. "wp-content" "gtag" = WordPress sites that also load Google Analytics.
a OR b  ·  a | bOR, match pages containing either term. PublicWWW can't do this; we can. "cdn.shopify.com" OR "cdn.bigcommerce.com".
-"exclude"  ·  NOT "exclude"NOT, a leading minus (or the uppercase keyword) removes pages that contain the term. "wp-content" -"woocommerce" = WordPress without WooCommerce.
AND   OR   NOTKeywords are operators only in uppercase. Lowercase and, or, not are ordinary words: terms and conditions searches for those three words, and "AND" in quotes is the literal string.
(  …  )Grouping to control precedence. Order is NOT > AND > OR, so parenthesize to mix them: ("a" OR "b") -"c". A parenthesis glued to a term is part of the literal, not a group: jQuery(document) and [].constructor.constructor(atob) are each one string, exactly as source code writes them. A group needs a space (or the start of the query) before its (, and a term that opens a parenthesis claims the ) that balances it: (jQuery(document) OR $(document)) -"noConflict" groups, while (a OR eval() is rejected as an unclosed group.
site:tldRestrict to a top-level domain. site:io, site:gov, site:com. (tld: is an alias.) It filters the top-level domain only, site:example.com is not a thing we can filter on, and says so rather than quietly finding nothing.
-site:tldExclude a top-level domain. "wp-content" -site:ru = WordPress sites, minus .ru. The minus binds to the filter, not to the next term. Several filters combine: site:com site:org is either TLD, -site:ru -site:cn excludes both, and the two kinds mix.
/regex/A regular expression over the source. /UA-\d{4,}-\d/ finds legacy Analytics IDs. Works standalone or as one term inside a boolean query. Both slashes are required, a bare path like /wp-content/themes is searched literally.
\"   \\Escaping, to search for a literal quote or backslash inside a phrase, escape it: "say \"hi\"". These are the only two escapes; any other backslash is part of the literal, so "https:\/\/cdn.example.com" finds the JSON-escaped form exactly as written.

Examples

Click any example to run it.

Coming from PublicWWW?

The core operators map 1-to-1, "phrase", space for AND, - to exclude, and /regex/ all behave the same. Two differences worth knowing:

Tips