From 65e3fa79b77524050d83af0d6b96f7caa6e90aba Mon Sep 17 00:00:00 2001
From: Benjamin Benoist <benjamin.benoist@snowplowanalytics.com>
Date: Tue, 3 Oct 2023 17:37:14 +0200
Subject: [PATCH] Handle /robots.txt

---
 .../Routes.scala                              |  3 +-
 .../scalastream/it/core/RobotsSpec.scala      | 65 +++++++++++++++++++
 2 files changed, 67 insertions(+), 1 deletion(-)
 create mode 100644 kinesis/src/it/scala/com/snowplowanalytics/snowplow/collectors/scalastream/it/core/RobotsSpec.scala

diff --git a/http4s/src/main/scala/com.snowplowanalytics.snowplow.collector.core/Routes.scala b/http4s/src/main/scala/com.snowplowanalytics.snowplow.collector.core/Routes.scala
index 3f9cad421..f03071c9b 100644
--- a/http4s/src/main/scala/com.snowplowanalytics.snowplow.collector.core/Routes.scala
+++ b/http4s/src/main/scala/com.snowplowanalytics.snowplow.collector.core/Routes.scala
@@ -21,7 +21,8 @@ class Routes[F[_]: Sync](enableDefaultRedirect: Boolean, service: IService[F]) e
           ifTrue  = Ok("OK"),
           ifFalse = ServiceUnavailable("Service Unavailable")
         )
-
+    case GET -> Root / "robots.txt" =>
+      Ok("User-agent: *\nDisallow: /")
   }
 
   private val corsRoute = HttpRoutes.of[F] {
diff --git a/kinesis/src/it/scala/com/snowplowanalytics/snowplow/collectors/scalastream/it/core/RobotsSpec.scala b/kinesis/src/it/scala/com/snowplowanalytics/snowplow/collectors/scalastream/it/core/RobotsSpec.scala
new file mode 100644
index 000000000..561c5220d
--- /dev/null
+++ b/kinesis/src/it/scala/com/snowplowanalytics/snowplow/collectors/scalastream/it/core/RobotsSpec.scala
@@ -0,0 +1,65 @@
+/*
+ * Copyright (c) 2023-2023 Snowplow Analytics Ltd. All rights reserved.
+ *
+ * This program is licensed to you under the Apache License Version 2.0, and
+ * you may not use this file except in compliance with the Apache License
+ * Version 2.0.  You may obtain a copy of the Apache License Version 2.0 at
+ * http://www.apache.org/licenses/LICENSE-2.0.
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the Apache License Version 2.0 is distributed on an "AS
+ * IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+ * implied.  See the Apache License Version 2.0 for the specific language
+ * governing permissions and limitations there under.
+ */
+package com.snowplowanalytics.snowplow.collectors.scalastream.it.core
+
+import scala.concurrent.duration._
+
+import org.specs2.mutable.Specification
+
+import cats.effect.IO
+
+import org.http4s.{Method, Request, Uri}
+
+import cats.effect.testing.specs2.CatsEffect
+
+import com.snowplowanalytics.snowplow.collectors.scalastream.it.kinesis.Kinesis
+import com.snowplowanalytics.snowplow.collectors.scalastream.it.kinesis.containers._
+import com.snowplowanalytics.snowplow.collectors.scalastream.it.Http
+
+class RobotsSpec extends Specification with Localstack with CatsEffect {
+
+  override protected val Timeout = 5.minutes
+
+  "collector" should {
+    "respond to /robots.txt with 200 and not emit any event" in {
+      val testName = "robots"
+      val streamGood = s"$testName-raw"
+      val streamBad = s"$testName-bad-1"
+
+      Collector.container(
+        "kinesis/src/it/resources/collector.hocon",
+        testName,
+        streamGood,
+        streamBad
+      ).use { collector =>
+        val uri = Uri.unsafeFromString(s"http://${collector.host}:${collector.port}/robots.txt")
+        val request = Request[IO](Method.GET, uri)
+
+        for {
+          response <- Http.response(request)
+          bodyBytes <- response.body.compile.toList
+          body = new String(bodyBytes.toArray)
+          _ <- IO.sleep(10.second)
+          collectorOutput <- Kinesis.readOutput(streamGood, streamBad)
+        } yield {
+          response.status.code must beEqualTo(200)
+          body must beEqualTo("User-agent: *\nDisallow: /")
+          collectorOutput.good must beEmpty
+          collectorOutput.bad must beEmpty
+        }
+      }
+    }
+  }
+}