//! Benchmark helper: read an HTML body from stdin, print the canonical
//! `engine_classify` tag + byte length. Lets the competitor benchmark
//! harness scores every browser's final
//! rendered DOM through the *exact same* classifier browser_oxide uses
//! for its own corpus sweep, so the comparison has zero classifier drift.
//!
//! Output: one line `<tag>\t<len>` on stdout.
use Read;