Merge pull request #2223 from DataDog/maciey/RUM-8931/host-sanitizer-fix

RUM-8931 Simplify host sanitizer logic

Author: maciejburda

DatadogInternal/Sources/NetworkInstrumentation/HostsSanitizer.swift

@@ -15,31 +15,21 @@ public protocol HostsSanitizing {
 }
 
 public struct HostsSanitizer: HostsSanitizing {
-    private let urlRegex = #"^(http|https)://(.*)"#
-    private let hostRegex = #"^(([a-zA-Z0-9]|[a-zA-Z0-9][a-zA-Z0-9-]*[a-zA-Z0-9])\.)+([A-Za-z]|[A-Za-z][A-Za-z0-9-]*[A-Za-z0-9])$"#
+    private let hostRegex = #"^(([a-zA-Z0-9]|[a-zA-Z0-9][a-zA-Z0-9\-]*[a-zA-Z0-9])\.)*([A-Za-z0-9]|[A-Za-z0-9][A-Za-z0-9\-]*[A-Za-z0-9])$"#
     private let ipRegex = #"^(([0-9]|[1-9][0-9]|1[0-9]{2}|2[0-4][0-9]|25[0-5])\.){3}([0-9]|[1-9][0-9]|1[0-9]{2}|2[0-4][0-9]|25[0-5])$"#
 
     public init() { }
 
     private func sanitize(host: String, warningMessage: String) -> (String?, String?) {
-        if host.range(of: urlRegex, options: .regularExpression) != nil {
+        if let sanitizedHost = URL(string: host)?.host {
             // if an URL is given instead of the host, take its `host` part
-            if let sanitizedHost = URL(string: host)?.host {
-                let warning = "'\(host)' is an url and will be sanitized to: '\(sanitizedHost)'."
-                return (sanitizedHost, warning)
-            } else {
-                // otherwise, drop
-                let warning = "'\(host)' is not a valid host name and will be dropped."
-                return (nil, warning)
-            }
-        } else if host.range(of: hostRegex, options: .regularExpression) != nil {
-            // if a valid host name is given, accept it
-            return (host, nil)
+            let warning = "'\(host)' is an url and will be sanitized to: '\(sanitizedHost)'."
+            return (sanitizedHost, warning)
         } else if host.range(of: ipRegex, options: .regularExpression) != nil {
             // if a valid IP address is given, accept it
             return (host, nil)
-        } else if host == "localhost" {
-            // if "localhost" given, accept it
+        } else if host.range(of: hostRegex, options: .regularExpression) != nil {
+            // if a valid host name is given, accept it
             return (host, nil)
         } else {
             // otherwise, drop

DatadogInternal/Tests/NetworkInstrumentation/HostsSanitizerTests.swift

@@ -21,31 +21,30 @@ class HostsSanitizerTests: XCTestCase {
             "https://first-party.com/v2/api", // sanitize to → "first-party.com"
             "https://192.168.0.1/api", // sanitize to → "192.168.0.1"
             "https://192.168.0.2", // sanitize to → "192.168.0.2"
-            "invalid-host-name", // drop
+            "invalid_host_name", // drop
             "192.168.0.3:8080", // drop
             "", // drop
             "localhost", // accept
             "192.168.0.4", // accept
             "valid-host-name.com", // accept
+            "customprotocol://name" // accept
         ]
 
         // Then
         let sanitizer = HostsSanitizer()
         let sanitizedHosts = sanitizer.sanitized(hosts: hosts, warningMessage: "Host is not valid")
 
-        XCTAssertEqual(
-            sanitizedHosts,
-            [
-                "first-party.com",
-                "api.first-party.com",
-                "localhost",
-                "192.168.0.1",
-                "192.168.0.2",
-                "localhost",
-                "192.168.0.4",
-                "valid-host-name.com"
-            ]
-        )
+        XCTAssertEqual(sanitizedHosts.count, 8)
+        XCTAssertTrue(sanitizedHosts.contains("first-party.com"))
+        XCTAssertTrue(sanitizedHosts.contains("api.first-party.com"))
+        XCTAssertTrue(sanitizedHosts.contains("192.168.0.1"))
+        XCTAssertTrue(sanitizedHosts.contains("192.168.0.2"))
+        XCTAssertTrue(sanitizedHosts.contains("localhost"))
+        XCTAssertTrue(sanitizedHosts.contains("192.168.0.4"))
+        XCTAssertTrue(sanitizedHosts.contains("valid-host-name.com"))
+        XCTAssertTrue(sanitizedHosts.contains("name"))
+
+        XCTAssertEqual(printFunction.printedMessages.count, 9)
 
         XCTAssertTrue(
             printFunction.printedMessages.contains("⚠️ Host is not valid: '192.168.0.3:8080' is not a valid host name and will be dropped.")
@@ -66,11 +65,13 @@ class HostsSanitizerTests: XCTestCase {
             printFunction.printedMessages.contains("⚠️ Host is not valid: 'https://first-party.com/v2/api' is an url and will be sanitized to: 'first-party.com'.")
         )
         XCTAssertTrue(
-            printFunction.printedMessages.contains("⚠️ Host is not valid: 'invalid-host-name' is not a valid host name and will be dropped.")
+            printFunction.printedMessages.contains("⚠️ Host is not valid: 'invalid_host_name' is not a valid host name and will be dropped.")
         )
         XCTAssertTrue(
             printFunction.printedMessages.contains("⚠️ Host is not valid: 'https://192.168.0.2' is an url and will be sanitized to: '192.168.0.2'.")
         )
-        XCTAssertEqual(printFunction.printedMessages.count, 8)
+        XCTAssertTrue(
+            printFunction.printedMessages.contains("⚠️ Host is not valid: 'customprotocol://name' is an url and will be sanitized to: 'name'.")
+        )
     }
 }