Merge #1534

1534: Require all traffic to provide a `User-Agent` header. r=sgrif a=sgrif

We want to be able to actually differentiate crawlers from each other,
so I've nudged them towards actually using a unique user agent (we
probably won't ever actually block generic UAs since folks sometimes do
actually use curl/wget from the command line).

Additionally, I've had a lot of cases lately where a crawler has been
outside of what we allow, but wasn't actually causing a service impact.
If I could contact those people without having to block their traffic, I
would. So I've also worded the message to try and nudge folks towards
including contact info, which most commercial bots already do.

Co-authored-by: Sean Griffin <[email protected]>
Co-authored-by: Justin Geibel <[email protected]>

Author: 3 people

src/middleware/mod.rs

@@ -20,6 +20,7 @@ mod ember_index_rewrite;
 mod ensure_well_formed_500;
 mod head;
 mod log_request;
+mod require_user_agent;
 mod security_headers;
 mod static_or_continue;
 
@@ -85,6 +86,7 @@ pub fn build_middleware(app: Arc<App>, endpoints: R404) -> MiddlewareBuilder {
         let ips = ip_list.split(',').map(String::from).collect();
         m.around(block_ips::BlockIps::new(ips));
     }
+    m.around(require_user_agent::RequireUserAgent::default());
 
     if env != Env::Test {
         m.around(log_request::LogRequests::default());

src/middleware/no_user_agent_message.txt

@@ -0,0 +1,13 @@
+We require that all requests include a `User-Agent` header.  To allow us to determine the impact your bot has on our service, we ask that your user agent actually identify your bot, and not just report the HTTP client library you're using.  Including contact information will also reduce the chance that we will need to take action against your bot.
+
+Bad:
+  User-Agent: reqwest/0.9.1
+
+Better:
+  User-Agent: my_crawler
+
+Best:
+  User-Agent: my_crawler (my_crawler.com/info)
+  User-Agent: my_crawler (help@my_crawler.com)
+
+If you believe you've received this message in error, please email [email protected] and include the request id {}.

src/middleware/require_user_agent.rs

@@ -0,0 +1,41 @@
+//! Middleware that blocks requests with no user-agent header
+
+use super::prelude::*;
+
+use std::collections::HashMap;
+use std::io::Cursor;
+use util::request_header;
+
+// Can't derive debug because of Handler.
+#[allow(missing_debug_implementations)]
+#[derive(Default)]
+pub struct RequireUserAgent {
+    handler: Option<Box<dyn Handler>>,
+}
+
+impl AroundMiddleware for RequireUserAgent {
+    fn with_handler(&mut self, handler: Box<dyn Handler>) {
+        self.handler = Some(handler);
+    }
+}
+
+impl Handler for RequireUserAgent {
+    fn call(&self, req: &mut dyn Request) -> Result<Response, Box<dyn Error + Send>> {
+        let has_user_agent = request_header(req, "User-Agent") != "";
+        if !has_user_agent {
+            let body = format!(
+                include_str!("no_user_agent_message.txt"),
+                request_header(req, "X-Request-Id"),
+            );
+            let mut headers = HashMap::new();
+            headers.insert("Content-Length".to_string(), vec![body.len().to_string()]);
+            Ok(Response {
+                status: (403, "Forbidden"),
+                headers,
+                body: Box::new(Cursor::new(body.into_bytes())),
+            })
+        } else {
+            self.handler.as_ref().unwrap().call(req)
+        }
+    }
+}

src/tests/all.rs

@@ -90,6 +90,7 @@ mod krate;
 mod owners;
 mod record;
 mod schema_details;
+mod server;
 mod team;
 mod token;
 mod user;
@@ -187,7 +188,9 @@ fn env(var: &str) -> String {
 }
 
 fn req(method: conduit::Method, path: &str) -> MockRequest {
-    MockRequest::new(method, path)
+    let mut request = MockRequest::new(method, path);
+    request.header("User-Agent", "conduit-test");
+    request
 }
 
 fn ok_resp(r: &conduit::Response) -> bool {

src/tests/server.rs

@@ -0,0 +1,13 @@
+use conduit::{Handler, Method};
+
+use {app, req};
+
+#[test]
+fn user_agent_is_required() {
+    let (_b, _app, middle) = app();
+
+    let mut req = req(Method::Get, "/api/v1/crates");
+    req.header("User-Agent", "");
+    let resp = t!(middle.call(&mut req));
+    assert_eq!(resp.status.0, 403);
+}

src/tests/util.rs

@@ -158,7 +158,7 @@ pub struct MockAnonymousUser {
 
 impl RequestHelper for MockAnonymousUser {
     fn request_builder(&self, method: Method, path: &str) -> MockRequest {
-        MockRequest::new(method, path)
+        ::req(method, path)
     }
 
     fn app(&self) -> &TestApp {
@@ -177,7 +177,7 @@ pub struct MockCookieUser {
 
 impl RequestHelper for MockCookieUser {
     fn request_builder(&self, method: Method, path: &str) -> MockRequest {
-        let mut request = MockRequest::new(method, path);
+        let mut request = ::req(method, path);
         request.mut_extensions().insert(self.user.clone());
         request
             .mut_extensions()
@@ -218,7 +218,7 @@ pub struct MockTokenUser {
 
 impl RequestHelper for MockTokenUser {
     fn request_builder(&self, method: Method, path: &str) -> MockRequest {
-        let mut request = MockRequest::new(method, path);
+        let mut request = ::req(method, path);
         request.header("Authorization", &self.token.token);
         request
     }