referer-parser-rs

Rust implementation of snowplow-referer-parser. Identifies the source of a HTTP referer URL (search engine, social network, webmail, etc.) and extracts the search term when available.

Usage

Download the referers.yml database, then:

use referer_parser_rs::Parser;
use url::Url;

let parser = Parser::new("referers.yml").unwrap();
let url = Url::parse("https://www.google.com/search?q=hello+world").unwrap();

if let Some(entry) = parser.lookup(&url) {
    println!("{} / {} / {:?}", entry.medium, entry.source, entry.search_term);
    // search / Google / Some("hello world")
}

Design

• The YAML database is loaded once into a flat Vec of entries, with a HashMap mapping each domain (or domain/path) key to an entry index.
• lookup returns an Entry that borrows from both the Parser and the input Url -- no heap allocation unless percent-decoding is needed for the search term.
• Domain resolution walks subdomains (www.google.fr -> google.fr) and path prefixes (/webmail/fr_FR -> /webmail), preferring path-qualified matches over domain-only matches.