rust_web_server/traffic/
mod.rs

1//! Traffic insights (18.15.0): who visits — people, AI agents, bots —
2//! worked out by rws from each request, for the admin UI's Traffic page.
3//!
4//! This first part is the classifier, [`classify`]: from the client's
5//! address, User-Agent and path it names what kind of client this is, using
6//! - the IP ranges Google, Bing, Apple, DuckDuckGo, OpenAI, Perplexity and
7//!   Common Crawl publish for their crawlers (a match is *verified*; a
8//!   User-Agent that claims one of them from elsewhere is *spoofed*);
9//! - [ai.robots.txt](https://github.com/ai-robots-txt/ai.robots.txt)'s AI
10//!   agents and [crawler-user-agents](https://github.com/monperrus/crawler-user-agents)'
11//!   bot patterns (both MIT, bundled, regenerated by `ci/gen_bot_lists.py`);
12//! - requests only scanners make (`/.env`, `/.git/config`, `/wp-login.php`…).
13//!
14//! A browser User-Agent that matches none of them is a [`Kind::Browser`]:
15//! whether it's a person is decided later, by whether that client loads the
16//! page's assets as a browser does.
17
18mod classify;
19pub mod config;
20mod data;
21pub(crate) mod persist;
22pub mod record;
23pub mod summary;
24
25pub use classify::{classify, Class, Group, Kind};
26pub use record::{enable, enabled};
27
28/// Count one finished request (called with the access log line, whatever
29/// the log format).
30pub(crate) fn observe(request: &crate::request::Request, response: &crate::response::Response, ip: std::net::IpAddr) {
31    if !enabled() {
32        return;
33    }
34    let header = |name: &str| request.header(name).map(|h| h.value.as_ref()).filter(|v: &&str| !v.is_empty());
35    let content_type = response
36        .content_range_list
37        .first()
38        .map(|c| c.content_type.as_str())
39        .or_else(|| response.headers.iter().find(|h| h.name.eq_ignore_ascii_case("content-type")).map(|h| h.value.as_ref()));
40    record::record(&record::Hit {
41        ip,
42        host: header("host").unwrap_or(""),
43        user_agent: header("user-agent"),
44        referer: header("referer"),
45        method: &request.method,
46        path: &request.request_uri,
47        status: response.status_code,
48        content_type,
49        now: record::now(),
50    });
51}
52
53/// Start counting, with the saved counts loaded and saved again every five
54/// minutes (when `persist` gives a file). The `rws` binary calls this when
55/// the admin API is on and `[traffic] enabled` isn't `false`.
56pub fn start() {
57    if enabled() {
58        return;
59    }
60    enable();
61    let Some(path) = config::persist_path() else { return };
62    persist::load(&path);
63    let _ = std::thread::Builder::new().name("rws-traffic".to_string()).spawn(move || loop {
64        std::thread::sleep(std::time::Duration::from_secs(300));
65        if let Err(e) = persist::save(&path) {
66            eprintln!("traffic: can't save {}: {e}", path.display());
67        }
68    });
69}
70
71/// Save the counts now (at shutdown).
72pub fn save() {
73    if let (true, Some(path)) = (enabled(), config::persist_path()) {
74        if let Err(e) = persist::save(&path) {
75            eprintln!("traffic: can't save {}: {e}", path.display());
76        }
77    }
78}
79
80#[cfg(test)]
81mod tests;
rust_web_server/traffic/mod.rs

rust_web_server/traffic/
mod.rs
