oscerd commented on code in PR #26734:
URL: https://github.com/apache/camel/pull/26734#discussion_r4080199333


##########
core/camel-base-engine/src/main/java/org/apache/camel/impl/engine/DefaultResourceResolvers.java:
##########
@@ -42,9 +43,47 @@
 
 public final class DefaultResourceResolvers {
 
+    /**
+     * Property key for how long to wait for the connection to an {@code 
http:} or {@code https:} resource to be
+     * established, in milliseconds. {@code 0} means wait indefinitely.
+     */
+    public static final String HTTP_CONNECT_TIMEOUT_PROPERTY = 
"camel.resource.http.connect-timeout";
+
+    /**
+     * Property key for how long to wait for data when reading an {@code 
http:} or {@code https:} resource, in
+     * milliseconds. {@code 0} means wait indefinitely.
+     */
+    public static final String HTTP_READ_TIMEOUT_PROPERTY = 
"camel.resource.http.read-timeout";
+
+    static final int DEFAULT_HTTP_CONNECT_TIMEOUT = 10000;
+    static final int DEFAULT_HTTP_READ_TIMEOUT = 30000;
+
     private DefaultResourceResolvers() {
     }
 
+    /**
+     * Resolves a timeout from the properties component, falling back to the 
given default.
+     * <p/>
+     * Resolution happens per resource rather than once per resolver because a 
resolver is a long-lived service while
+     * the properties it reads can be reloaded underneath it.
+     */
+    private static int resolveTimeout(CamelContext camelContext, String key, 
int defaultValue) {
+        if (camelContext == null) {
+            return defaultValue;
+        }
+        return camelContext.getPropertiesComponent()
+                .resolveProperty(key)
+                .map(Integer::parseInt)
+                .orElse(defaultValue);

Review Comment:
   Applied in `f012ca85`, including the `trim()`.
   
   You are right and my reasoning for the original was weak. I had deliberately 
matched `OAuthTokenValidationConfigResolver`, which parses `camel.oauth.*` the 
same bare way, on the grounds that following an existing pattern beats 
inventing a nicer one. But that resolver runs behind a feature an operator 
opted into, whereas this one runs on any context that resolves an `http:` 
resource — so the population who can hit it is much larger, and they hit it at 
startup with no clue which key is at fault.
   
   Added a test for it, asserting on the **message** rather than the type. That 
distinction matters here: `NumberFormatException extends 
IllegalArgumentException`, so `assertThrows(IllegalArgumentException.class, 
...)` passes with or without the fix. Reverting the catch reddens it with:
   
   ```
   message should name the property, was: For input string: "10s"
   ```
   
   which is precisely the message this change exists to replace.
   
   _Claude Code on behalf of @oscerd_



##########
core/camel-core/src/test/java/org/apache/camel/util/HttpResourceTimeoutTest.java:
##########
@@ -0,0 +1,173 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.camel.util;
+
+import java.io.IOException;
+import java.net.ServerSocket;
+import java.net.Socket;
+import java.net.SocketTimeoutException;
+import java.util.Properties;
+import java.util.concurrent.CountDownLatch;
+import java.util.concurrent.TimeUnit;
+
+import org.apache.camel.CamelContext;
+import org.apache.camel.impl.DefaultCamelContext;
+import org.apache.camel.impl.engine.DefaultResourceResolvers;
+import org.apache.camel.spi.Resource;
+import org.apache.camel.support.ResourceHelper;
+import org.junit.jupiter.api.AfterEach;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+import org.junit.jupiter.api.Timeout;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertFalse;
+import static org.junit.jupiter.api.Assertions.assertThrows;
+import static org.junit.jupiter.api.Assertions.assertTrue;
+
+/**
+ * An {@code http:} resource must not be able to stall the caller forever.
+ * <p/>
+ * The server here accepts the connection and then never answers, which is the 
only shape that reproduces the problem: a
+ * refused connection fails fast on its own, so nothing is proven by pointing 
at a closed port.
+ * <p/>
+ * The timeouts below are deliberately {@code SEPARATE_THREAD}: the default 
mode only measures elapsed time once the
+ * test method returns, so a regression that restores the indefinite wait 
would hang the build instead of failing it.
+ */
+public class HttpResourceTimeoutTest {
+
+    private ServerSocket server;
+    private Thread acceptor;
+    private CountDownLatch accepted;
+    private volatile boolean stopped;
+
+    @BeforeEach
+    void startMuteServer() throws Exception {
+        server = new ServerSocket(0);
+        accepted = new CountDownLatch(1);
+        acceptor = new Thread(() -> {
+            while (!stopped) {
+                try (Socket socket = server.accept()) {
+                    accepted.countDown();
+                    // hold the connection open and write nothing at all, so 
the client is left waiting on a read
+                    while (!stopped && !socket.isClosed()) {
+                        Thread.onSpinWait();
+                    }

Review Comment:
   Applied. Good catch — this was a real defect in the harness, not a style 
point.
   
   `socket.isClosed()` only reflects a **local** close, so the loop could never 
observe the client going away; it spun a core for the full duration of every 
mute-server test and only exited when `@AfterEach` set `stopped`. 
`transferTo(OutputStream.nullOutputStream())` blocks on the read instead and 
returns when the client gives up, which is the same observable behaviour for 
the client at no CPU cost.
   
   _Claude Code on behalf of @oscerd_



##########
core/camel-core/src/test/java/org/apache/camel/util/HttpResourceTimeoutTest.java:
##########
@@ -0,0 +1,173 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.camel.util;
+
+import java.io.IOException;
+import java.net.ServerSocket;
+import java.net.Socket;
+import java.net.SocketTimeoutException;
+import java.util.Properties;
+import java.util.concurrent.CountDownLatch;
+import java.util.concurrent.TimeUnit;
+
+import org.apache.camel.CamelContext;
+import org.apache.camel.impl.DefaultCamelContext;
+import org.apache.camel.impl.engine.DefaultResourceResolvers;
+import org.apache.camel.spi.Resource;
+import org.apache.camel.support.ResourceHelper;
+import org.junit.jupiter.api.AfterEach;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+import org.junit.jupiter.api.Timeout;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertFalse;
+import static org.junit.jupiter.api.Assertions.assertThrows;
+import static org.junit.jupiter.api.Assertions.assertTrue;
+
+/**
+ * An {@code http:} resource must not be able to stall the caller forever.
+ * <p/>
+ * The server here accepts the connection and then never answers, which is the 
only shape that reproduces the problem: a
+ * refused connection fails fast on its own, so nothing is proven by pointing 
at a closed port.
+ * <p/>
+ * The timeouts below are deliberately {@code SEPARATE_THREAD}: the default 
mode only measures elapsed time once the
+ * test method returns, so a regression that restores the indefinite wait 
would hang the build instead of failing it.
+ */
+public class HttpResourceTimeoutTest {
+
+    private ServerSocket server;
+    private Thread acceptor;
+    private CountDownLatch accepted;
+    private volatile boolean stopped;
+
+    @BeforeEach
+    void startMuteServer() throws Exception {
+        server = new ServerSocket(0);
+        accepted = new CountDownLatch(1);
+        acceptor = new Thread(() -> {
+            while (!stopped) {
+                try (Socket socket = server.accept()) {
+                    accepted.countDown();
+                    // hold the connection open and write nothing at all, so 
the client is left waiting on a read
+                    while (!stopped && !socket.isClosed()) {
+                        Thread.onSpinWait();
+                    }
+                } catch (IOException e) {
+                    return;
+                }
+            }
+        }, "mute-http-server");
+        acceptor.setDaemon(true);
+        acceptor.start();
+        stopped = false;
+    }
+
+    @AfterEach
+    void stopMuteServer() throws Exception {
+        stopped = true;
+        server.close();
+        acceptor.join(TimeUnit.SECONDS.toMillis(5));
+    }
+
+    private CamelContext contextWithReadTimeout(String millis) {
+        CamelContext context = new DefaultCamelContext();
+        Properties properties = new Properties();
+        
properties.setProperty(DefaultResourceResolvers.HTTP_READ_TIMEOUT_PROPERTY, 
millis);
+        context.getPropertiesComponent().setInitialProperties(properties);
+        context.start();
+        return context;
+    }
+
+    private String muteUrl() {
+        return "http://localhost:"; + server.getLocalPort() + "/policy.rego";
+    }
+
+    @Test
+    @Timeout(value = 30, threadMode = Timeout.ThreadMode.SEPARATE_THREAD)
+    public void getInputStreamGivesUpOnAServerThatNeverAnswers() throws 
Exception {
+        CamelContext context = contextWithReadTimeout("500");
+        try {
+            Resource resource = ResourceHelper.resolveResource(context, 
muteUrl());
+
+            assertThrows(SocketTimeoutException.class, 
resource::getInputStream);
+            assertTrue(accepted.await(5, TimeUnit.SECONDS),
+                    "the server should have accepted the connection - a 
refused connect proves nothing");
+        } finally {
+            context.stop();
+        }
+    }
+
+    @Test
+    @Timeout(value = 30, threadMode = Timeout.ThreadMode.SEPARATE_THREAD)
+    public void existsGivesUpOnAServerThatNeverAnswers() throws Exception {
+        CamelContext context = contextWithReadTimeout("500");
+        try {
+            Resource resource = ResourceHelper.resolveResource(context, 
muteUrl());
+
+            // exists() wraps the IOException rather than declaring it, so the 
timeout surfaces as the cause
+            IllegalArgumentException e = 
assertThrows(IllegalArgumentException.class, resource::exists);
+            assertEquals(SocketTimeoutException.class, 
e.getCause().getClass());
+        } finally {
+            context.stop();
+        }
+    }
+
+    @Test
+    @Timeout(value = 30, threadMode = Timeout.ThreadMode.SEPARATE_THREAD)
+    public void httpsResourcesAreBoundedToo() throws Exception {
+        CamelContext context = contextWithReadTimeout("500");
+        try {
+            // the https resolver hands back the same HttpResource, so it must 
inherit the same bound; the handshake
+            // against a mute plain-text server is what fails here, and it 
must fail rather than hang
+            Resource resource
+                    = ResourceHelper.resolveResource(context, 
"https://localhost:"; + server.getLocalPort() + "/x");
+
+            assertThrows(IOException.class, resource::getInputStream);
+        } finally {
+            context.stop();
+        }
+    }
+
+    @Test
+    @Timeout(value = 30, threadMode = Timeout.ThreadMode.SEPARATE_THREAD)
+    public void aResourceThatIsNotThereStillReportsAbsent() throws Exception {
+        // guards against the timeouts turning an ordinary 404 into a failure
+        CamelContext context = contextWithReadTimeout("5000");
+        try (ServerSocket notFound = new ServerSocket(0)) {
+            Thread responder = new Thread(() -> {
+                try (Socket socket = notFound.accept()) {
+                    socket.getOutputStream()
+                            .write("HTTP/1.1 404 Not Found\r\nContent-Length: 
0\r\nConnection: close\r\n\r\n"
+                                    .getBytes("US-ASCII"));

Review Comment:
   Applied — `StandardCharsets.US_ASCII`, and it drops the 
`UnsupportedEncodingException` path as you note.
   
   _Claude Code on behalf of @oscerd_



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to