New WSDL Parser#2870
Merged
Merged
Conversation
… handling, and upgrade dependencies - Consolidated and corrected static imports for consistency across classes. - Unified handling of URI factories and base locations for WSDL and schema validation to reduce redundancy. - Replaced redundant code and removed deprecated methods for improved maintainability. - Upgraded dependencies: Jackson to 2.21.1, added jackson-bom, and updated Maven dependency configurations. - Deprecated unused test cases and streamlined the test suite.
Contributor
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
📝 WalkthroughWalkthroughThis pull request introduces a comprehensive WSDL parser library and refactors the validation and routing infrastructure to use it. Changes include new WSDL parsing classes, updates to schema validation, interceptor rewiring for path-based WSDL handling, dependency version updates, and test resources for WSDL scenarios. Changes
Sequence Diagram(s)sequenceDiagram
participant Client
participant SOAPProxy
participant WSDLParser as Definitions.parse()
participant WSDLInterceptor
participant ValidatorInterceptor
participant WSDLSchemaExtractor
Client->>SOAPProxy: init()
SOAPProxy->>WSDLParser: parse(resolver, wsdl)
WSDLParser->>WSDLSchemaExtractor: extract embedded schemas
WSDLSchemaExtractor-->>WSDLParser: List<Document>
WSDLParser-->>SOAPProxy: Definitions
SOAPProxy->>WSDLInterceptor: setSoapProxy(this)
SOAPProxy->>ValidatorInterceptor: setSoapProxy(this)
SOAPProxy->>WSDLInterceptor: init()
WSDLInterceptor->>WSDLInterceptor: setPathRewriterOnWSDLInterceptor(path)
WSDLInterceptor-->>SOAPProxy: initialized
Estimated code review effort🎯 4 (Complex) | ⏱️ ~50 minutes Possibly related PRs
Suggested labels
Suggested reviewers
Poem
✨ Finishing Touches🧪 Generate unit tests (beta)
|
Member
Author
|
@coderabbitai review |
Contributor
✅ Actions performedReview triggered.
|
Member
Author
|
@coderabbitai resume |
- Ensured all relevant Java files across core and annot modules include the proper Apache License 2.0 headers.
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 19
Note
Due to the large number of review comments, Critical, Major severity comments were prioritized as inline comments.
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
annot/pom.xml (1)
48-63:⚠️ Potential issue | 🟠 MajorKeep the
annotLog4j dependencies test-scoped.The
log4j-slf4j2-implandlog4j-coredependencies inannot/pom.xmlare missing<scope>test</scope>. They should be scoped to tests in the annot module while remaining compile-scoped only in the distribution module.Suggested fix
<dependency> <groupId>org.apache.logging.log4j</groupId> <artifactId>log4j-slf4j2-impl</artifactId> + <scope>test</scope> </dependency> <dependency> <groupId>org.apache.logging.log4j</groupId> <artifactId>log4j-core</artifactId> + <scope>test</scope> </dependency>🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@annot/pom.xml` around lines 48 - 63, The Log4j dependencies log4j-slf4j2-impl and log4j-core in the annot module are missing test scope: update the dependency entries for artifactId "log4j-slf4j2-impl" and "log4j-core" to include <scope>test</scope> so they are only used for tests in the annot module (leave other dependencies like jackson-databind and jackson-annotations unchanged); ensure you modify the dependency declarations for those two artifactIds in the pom so they are test-scoped.core/src/main/java/com/predic8/membrane/core/ws/relocator/Relocator.java (1)
43-45:⚠️ Potential issue | 🟠 MajorKeep the public constructor accepting
Writer.Nothing in this implementation requires
OutputStreamWriter. The change narrows the public API for no functional gain and breaks callers that useStringWriteror customWriterimplementations. The commented-out test inRelocatorWADLTest.java(line 60) shows this was previously supported. For 6.X, preserve the wider type—XMLOutputFactory.createXMLEventWriter()acceptsWriternatively.♻️ Proposed fix
- public Relocator(OutputStreamWriter osw, String protocol, String host, + public Relocator(Writer out, String protocol, String host, int port, String contextPath, PathRewriter pathRewriter) throws Exception { - this.writer = XMLOutputFactory.newInstance().createXMLEventWriter(osw); + this.writer = XMLOutputFactory.newInstance().createXMLEventWriter(out);🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/ws/relocator/Relocator.java` around lines 43 - 45, The public constructor currently narrows the API by taking OutputStreamWriter; revert it to accept java.io.Writer instead (change Relocator(OutputStreamWriter osw, String protocol, String host, int port, String contextPath, PathRewriter pathRewriter) to Relocator(Writer writer, ...)), and pass that Writer into XMLOutputFactory.newInstance().createXMLEventWriter(writer) so StringWriter and other Writer implementations continue to work (ensure signature visibility remains public and update any direct references/tests such as RelocatorWADLTest).
🟡 Minor comments (6)
core/src/test/resources/ws/import/message-references-import.wsdl-37-38 (1)
37-38:⚠️ Potential issue | 🟡 MinorBinding operation name mismatch with portType.
The binding operation is named
"test"(line 37) but the portType operation is named"get"(line 28). WSDL binding operations must reference valid portType operations by name.If this mismatch is intentional for testing parser validation, please disregard. Otherwise:
Proposed fix
- <operation name="test"> - <soap:operation soapAction="test"/> + <operation name="get"> + <soap:operation soapAction="get"/>🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/test/resources/ws/import/message-references-import.wsdl` around lines 37 - 38, The binding declares an operation named "test" but the portType defines the operation as "get", causing a name mismatch; update the binding <operation name="test"> to use the portType operation name "get" (or alternatively rename the portType operation to "test" if that was intended) and ensure the corresponding soap:operation soapAction is updated to match the chosen operation name so the binding correctly references the portType operation.core/src/test/resources/ws/import/message-references-import.wsdl-23-25 (1)
23-25:⚠️ Potential issue | 🟡 MinorUndefined element reference
m:Response.The
GetResponsemessage referencesm:Response, but the importedmessages.xsddefinesm:getResponse, notm:Response. This will cause schema validation to fail.If this is intentional for testing error scenarios, please disregard. Otherwise, the element reference should be corrected.
Proposed fix
<message name="GetResponse"> - <part name="parameters" element="m:Response"/> + <part name="parameters" element="m:getResponse"/> </message>🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/test/resources/ws/import/message-references-import.wsdl` around lines 23 - 25, The WSDL message GetResponse references a non-existent element m:Response; update the part element reference in the GetResponse message to the correct element defined in messages.xsd (m:getResponse) so the part attribute uses the existing element name (modify the <message name="GetResponse"> part element reference from m:Response to m:getResponse).core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/ProtocolOperation.java-1-16 (1)
1-16:⚠️ Potential issue | 🟡 MinorMissing license header.
This file is missing the Apache 2.0 license header that is present in other new files added in this PR (e.g.,
Address.java,OperationTest.java).📄 Proposed fix to add license header
+/* Copyright 2026 predic8 GmbH, www.predic8.com + + Licensed 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 com.predic8.membrane.core.util.wsdl.parser;🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/ProtocolOperation.java` around lines 1 - 16, The file ProtocolOperation.java is missing the Apache 2.0 license header; add the same standard Apache 2.0 header used in other new files (e.g., Address.java, OperationTest.java) at the top of the file before the package declaration so the class ProtocolOperation and its constructor/methods (ProtocolOperation(WSDLParserContext, Node) and getSoapAction()) have the required license comment header.core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Part.java-28-37 (1)
28-37:⚠️ Potential issue | 🟡 MinorAdd
@Nullableannotations to public QName accessor methods.Lines 28 and 32 declare non-nullable
QNamereturn types but delegate to a method marked@Nullableat line 36. The public API contract should align with the implementation.Recommended fix
- public QName getTypeQName() { + public `@Nullable` QName getTypeQName() { return getAttributeQName("type"); } - public QName getElementQName() { + public `@Nullable` QName getElementQName() { return getAttributeQName("element"); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Part.java` around lines 28 - 37, The public accessors getTypeQName and getElementQName currently declare non-nullable QName returns but delegate to the `@Nullable` getAttributeQName; update their signatures to return `@Nullable` QName (i.e. annotate both getTypeQName and getElementQName with the same `@Nullable` used on getAttributeQName) so the public API matches implementation, and add any required import for the chosen `@Nullable` annotation if missing.core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Port.java-27-29 (1)
27-29:⚠️ Potential issue | 🟡 Minor
getFirst()may throwNoSuchElementExceptionif no address element exists.If the WSDL port lacks an
<address>child element (e.g., abstract WSDL or malformed document),instantiateElements(...).getFirst()will throw an unchecked exception. Consider defensive handling similar togetBinding().🛡️ Proposed defensive handling
public Address getAddress() { - return instantiateElements(element,"address",Address.class).getFirst(); + var addresses = instantiateElements(element, "address", Address.class); + if (addresses.isEmpty()) { + throw new WSDLParserException("No address found for port: " + getName()); + } + return addresses.getFirst(); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Port.java` around lines 27 - 29, The getAddress() method currently calls instantiateElements(element, "address", Address.class).getFirst() which can throw NoSuchElementException when no <address> child exists; update getAddress() to defensively handle the empty case like getBinding() does by checking the instantiated list for emptiness and returning null (or an appropriate absent value) instead of calling getFirst() directly so missing address elements don't cause unchecked exceptions.core/src/main/java/com/predic8/membrane/core/interceptor/server/WSDLPublisherInterceptor.java-178-184 (1)
178-184:⚠️ Potential issue | 🟡 MinorConsider defensive null-check for
soapProxy.getWsdl().If
soapProxy.getWsdl()returns null,setWsdl(null)will be called at line 183, which clears all paths and adds null todocumentsToProcess. This could cause issues downstream.🛡️ Proposed defensive check
private void getWSDLFromEmbeddingSOAPProxy() { if (soapProxy == null) { throw new ConfigurationException("<wsdlPublisher> can only be used within a <soapProxy> or needs to declare <wsdlPublisher wsdl='...'>"); } wsdl = soapProxy.getWsdl(); + if (wsdl == null) { + throw new ConfigurationException("<wsdlPublisher> requires <soapProxy wsdl='...'> to be set"); + } setWsdl(wsdl); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/interceptor/server/WSDLPublisherInterceptor.java` around lines 178 - 184, getWSDLFromEmbeddingSOAPProxy() currently calls soapProxy.getWsdl() and unconditionally passes the result to setWsdl(), which can clear paths and add null into documentsToProcess; add a defensive null-check: after obtaining wsdl = soapProxy.getWsdl(), if wsdl is null throw a ConfigurationException (or log and return) with a clear message that the embedding <soapProxy> has no WSDL, otherwise call setWsdl(wsdl). Update the method to reference soapProxy and getWsdl() and ensure setWsdl() is only called with a non-null wsdl to avoid corrupting documentsToProcess.
🧹 Nitpick comments (12)
core/src/main/java/com/predic8/membrane/core/util/soap/WSDLUtil.java (1)
27-30: Consider adding null-safety for public API.Both
pathandreplacementNameparameters could cause issues if null:matcher(null)throwsNullPointerException, andformatted(null)embeds the literal string "null" in the output.🛡️ Optional: Add null checks
public static String rewriteRelativeWsdlPath(String path, String replacementName) { + if (path == null || replacementName == null) { + throw new IllegalArgumentException("path and replacementName must not be null"); + } return relativePathPattern .matcher(path) .replaceAll(quoteReplacement("./%s?".formatted(replacementName))); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/util/soap/WSDLUtil.java` around lines 27 - 30, rewriteRelativeWsdlPath currently calls matcher(path) and formatted(replacementName) without guarding against nulls; add null-safety at the start of the public method by validating inputs (e.g., use Objects.requireNonNull(path, "path must not be null") and either requireNonNull(replacementName, "replacementName must not be null") or coerce replacementName to "" when null) so matcher(...) never receives null and formatted(...) does not produce the literal "null"; keep the rest of the method (relativePathPattern.matcher(...).replaceAll(quoteReplacement(...))) unchanged and throw clear IllegalArgumentException/NullPointerException messages if you choose to reject nulls.core/src/test/java/com/predic8/membrane/core/interceptor/schemavalidation/SOAPUtilTest.java (1)
127-129: Prefer classpath resource loading over filesystem-relative paths in tests.Line 127–129 makes tests dependent on working directory layout. Loading from classpath is more stable across IDE/Maven/CI runs.
♻️ Proposed refactor
- private Message getMessage(String path) throws Exception { - return ok().contentType(TEXT_XML).body(new FileInputStream(path), true).build(); - } + private Message getMessage(String resourcePath) throws IOException { + InputStream in = SOAPUtilTest.class.getClassLoader().getResourceAsStream(resourcePath); + assertNotNull(in, "Missing test resource: " + resourcePath); + return ok().contentType(TEXT_XML).body(in, true).build(); + }Also update call sites to classpath-relative values, e.g.:
wsdlValidator/soapFaultFromSpec.xmlwsdlValidator/soapFaultCustom.xml🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/test/java/com/predic8/membrane/core/interceptor/schemavalidation/SOAPUtilTest.java` around lines 127 - 129, The test's getMessage method currently opens files via new FileInputStream(path) which makes tests reliant on the working directory; change getMessage to load the test file from the classpath (e.g., use getClass().getResourceAsStream(path) or Thread.currentThread().getContextClassLoader().getResourceAsStream(path)) and pass that InputStream into ok().contentType(TEXT_XML).body(...).build(); then update all call sites that pass filesystem-relative strings to use classpath-relative resource paths such as "wsdlValidator/soapFaultFromSpec.xml" and "wsdlValidator/soapFaultCustom.xml" so resources are resolved consistently in IDE/Maven/CI.core/src/test/resources/ws/include/xsd/inc/MessageStructureTypes2.xsd (1)
1-8: Consider using the imported namespace induto strengthen parser coverage.Right now the import is present but unused; adding one
dt:*reference makes this fixture validate cross-namespace resolution end-to-end.Proposed test-fixture refinement
-<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" +<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" + xmlns:dt="http://example.com/test/data-types" targetNamespace="http://example.com/test" elementFormDefault="qualified"> @@ - <xs:element name="du"/> + <xs:element name="du"> + <xs:complexType> + <xs:sequence> + <xs:element ref="dt:B5"/> + </xs:sequence> + </xs:complexType> + </xs:element>🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/test/resources/ws/include/xsd/inc/MessageStructureTypes2.xsd` around lines 1 - 8, The XSD imports the data-types namespace but never uses it; update MessageStructureTypes2.xsd so the imported namespace is referenced by declaring a prefix (e.g., xmlns:dt="http://example.com/test/data-types") and change the xs:element named "du" to reference a type or element from that namespace (for example set du to type="dt:SomeType" or include a child element from the dt: namespace) so cross-namespace resolution is exercised; locate the xs:element name="du" and the xs:import statement and add the xmlns:dt declaration and the dt: reference there.core/src/test/resources/ws/include/include.xsd (1)
6-16: Optional cleanup: remove commented-out schema directives in fixture.Keeping only active declarations makes this resource easier to read and maintain.
Suggested cleanup
-<!-- <xs:include schemaLocation="./xsd/messages.xsd"/>--> <xs:include schemaLocation="xsd/inc/MessageStructureTypes2.xsd"/> -<!-- <xs:import namespace="http://example.com/test/messages" schemaLocation="./xsd/messages.xsd"/>--> @@ - <xs:element name="test"> - </xs:element> + <xs:element name="test"/> @@ - <xs:element name="testResponse"> - </xs:element> + <xs:element name="testResponse"/>🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/test/resources/ws/include/include.xsd` around lines 6 - 16, Remove the commented-out schema directives to clean up the fixture: delete the commented lines containing "<!-- <xs:include schemaLocation="./xsd/messages.xsd"/> -->" and "<!-- <xs:import namespace="http://example.com/test/messages" schemaLocation="./xsd/messages.xsd"/> -->" so only the active xs:include (schemaLocation="xsd/inc/MessageStructureTypes2.xsd") and the xs:element definitions "test" and "testResponse" remain, improving readability and maintainability.pom.xml (1)
273-287: Align the XMLUnit module versions.
xmlunit-coreis now managed at 2.10.0, butxmlunit-matchersstays at 2.9.1.xmlunit-matchers2.10.0 is published, and that release targetsxmlunit-core2.10.0, so this leaves the parent POM with an avoidable mixed XMLUnit test stack. (repo.maven.apache.org)Suggested fix
<dependency> <groupId>org.xmlunit</groupId> <artifactId>xmlunit-matchers</artifactId> - <version>2.9.1</version> + <version>2.10.0</version> </dependency>🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@pom.xml` around lines 273 - 287, The POM has mixed XMLUnit versions: xmlunit-core is 2.10.0 while xmlunit-matchers remains 2.9.1; update the xmlunit-matchers dependency entry (artifactId: xmlunit-matchers) to use version 2.10.0 so both modules align with xmlunit-core, then run the build/tests to verify compatibility.core/src/test/java/com/predic8/membrane/core/interceptor/schemavalidation/SOAPFaultTest.java (1)
39-39: LGTM with minor style inconsistency.The
varusage is fine, but note thattestSkipFaults(line 47) andtestSkipFault2(line 55) still use explicitValidatorInterceptortype. Consider applyingvarconsistently across all test methods if that's the intended style direction.,
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/test/java/com/predic8/membrane/core/interceptor/schemavalidation/SOAPFaultTest.java` at line 39, Replace the explicit local variable type with "var" in the test methods that still declare a ValidatorInterceptor explicitly (look for local variables named like the interceptor in testSkipFaults and testSkipFault2) so that they match the style used when assigning createValidatorInterceptor(false); i.e., change "ValidatorInterceptor i = ..." to "var i = ..." in those test methods.core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/WSDLParserException.java (1)
17-20: Add acauseconstructor to preserve parse failure context.Current API only supports message text; wrapping underlying exceptions will lose stack provenance unless callers chain manually.
♻️ Suggested update
public class WSDLParserException extends RuntimeException { public WSDLParserException(String message) { super(message); } + + public WSDLParserException(String message, Throwable cause) { + super(message, cause); + } }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/WSDLParserException.java` around lines 17 - 20, WSDLParserException currently only has a String constructor so underlying parse exceptions lose stack/context; add overloaded constructors to preserve causes: implement WSDLParserException(String message, Throwable cause) that calls super(message, cause) and optionally WSDLParserException(Throwable cause) that calls super(cause) (and keep the existing WSDLParserException(String) for compatibility) so callers can wrap/paraphrase parse failures without losing the original exception.core/src/test/java/com/predic8/membrane/core/ws/WSDLIncludeImportTest.java (1)
102-104: Prefer explicit null-check for test resource lookup.
Line 103can throw an unhelpful NPE if the resource path changes;Objects.requireNonNull(...)gives clearer failure diagnostics.Suggested test-hardening diff
- var validator = new WSDLValidator(new ResolverMap(), - WSDLIncludeImportTest.class.getResource("/ws/include/include.wsdl").toString(), + var validator = new WSDLValidator(new ResolverMap(), + Objects.requireNonNull( + WSDLIncludeImportTest.class.getResource("/ws/include/include.wsdl"), + "Missing test resource /ws/include/include.wsdl" + ).toString(), "TestService", (msg, exc) -> log.info("Validation failure: {}", msg), true);🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/test/java/com/predic8/membrane/core/ws/WSDLIncludeImportTest.java` around lines 102 - 104, Wrap the test resource lookup in an explicit null-check using Objects.requireNonNull to avoid an ambiguous NPE: call Objects.requireNonNull(WSDLIncludeImportTest.class.getResource("/ws/include/include.wsdl"), "resource '/ws/include/include.wsdl' not found") and then call toString() for the WSDLValidator constructor; also add the java.util.Objects import if missing. This change targets the WSDLIncludeImportTest test where WSDLValidator(...) is constructed.core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/schema/Import.java (1)
52-62: Consider consolidating duplicate empty-check logic.
hasSchemaLocation()(lines 52-54) and the condition ingetSchema()(line 58) perform the same check. You could reusehasSchemaLocation()in the override.♻️ DRY improvement
`@Override` protected Schema getSchema(WSDLParserContext ctx) { - if (schemaLocation == null || schemaLocation.isEmpty()) { + if (!hasSchemaLocation()) { return null; } return super.getSchema(ctx); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/schema/Import.java` around lines 52 - 62, The null/empty check for schemaLocation is duplicated; update getSchema(WSDLParserContext ctx) to call the existing hasSchemaLocation() instead of repeating the condition: if hasSchemaLocation() returns false, return null, otherwise delegate to super.getSchema(ctx). Keep the existing hasSchemaLocation() method as the single source of truth for the empty-check so future changes only need to be made in one place.core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/schema/AbstractIncludeImport.java (1)
49-51: Wrap exceptions with more context for debugging.Rethrowing a generic
RuntimeExceptionloses information about which schema location failed to resolve. Consider including theschemaLocationin the exception message.♻️ Proposed improvement
} catch (Exception e) { - throw new RuntimeException(e); + throw new RuntimeException("Failed to resolve schema at location: " + schemaLocation, e); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/schema/AbstractIncludeImport.java` around lines 49 - 51, The catch block in AbstractIncludeImport currently rethrows a generic RuntimeException losing context; update the catch to throw a new RuntimeException that includes the schemaLocation (and any identifying info) in the message and pass the original exception as the cause (e.g., throw new RuntimeException("Failed to resolve schemaLocation: "+ schemaLocation, e)); this keeps the original stacktrace and makes it clear which schema failed to resolve.core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/BindingStyle.java (1)
1-6: Missing copyright header.Other new files in this PR include the Apache 2.0 license header. This file should have the same header for consistency.
📝 Add license header
+/* Copyright 2026 predic8 GmbH, www.predic8.com + + Licensed 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 com.predic8.membrane.core.util.wsdl.parser;🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/BindingStyle.java` around lines 1 - 6, Add the standard Apache 2.0 license header to the top of BindingStyle.java (the file in package com.predic8.membrane.core.util.wsdl.parser) so it matches other new files in the PR; insert the same multi-line Apache 2.0 copyright/license block used elsewhere in the repo above the package declaration in BindingStyle.java.core/src/main/java/com/predic8/membrane/core/interceptor/schemavalidation/ValidatorInterceptor.java (1)
153-166: Keep context on these error logs.Both branches log with an empty message, which makes search, aggregation, and alerting much harder in production. A short static prefix is enough.
💡 Suggested change
- log.error("", e); + log.error("Could not read message body during validation.", e); @@ - log.error("", e); + log.error("Validator execution failed.", e);🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@core/src/main/java/com/predic8/membrane/core/interceptor/schemavalidation/ValidatorInterceptor.java` around lines 153 - 166, The empty log.error calls in the ReadingBodyException catch and the generic Exception catch around validator.validateMessage remove context; change them to include a short static prefix (e.g. "ValidatorInterceptor: could not read message body" in the ReadingBodyException block and "ValidatorInterceptor: validation failed" in the validator.validateMessage catch) while still passing the exception object to log.error; update the log.error invocations near the ReadingBodyException handler and the catch(Exception e) after validator.validateMessage() to use these descriptive messages so the internal(...).detail(...).exception(e) flow remains unchanged.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In
`@core/src/main/java/com/predic8/membrane/core/interceptor/schemavalidation/AbstractXMLSchemaValidator.java`:
- Around line 68-74: The init() loop currently calls createValidators()
repeatedly which rebuilds SchemaFactory and recompiles all schemas for each pool
entry; change initialization to compile and cache Schema objects once (using
SchemaFactory.newSchema(source)) and then populate the validators queue by
creating fresh Validator instances from the cached Schema(s)
(Schema.newValidator()) instead of recompiling; update createValidators() (or
add a helper) so it reuses the cached Schema instances to produce Validators,
and ensure any SchemaFactory/Scheme compilation logic lives in a single
initialization step and that validators is filled with Validators created from
those cached Schema objects.
In
`@core/src/main/java/com/predic8/membrane/core/interceptor/schemavalidation/WSDLMessageElementExtractor.java`:
- Around line 61-65: The current early-return in getTypesByStyle suppresses
serviceName-based validation and style detection when Definitions.getServices()
is empty, forcing all PortTypes into the document bucket; change getTypesByStyle
to not short-circuit on empty services but instead call
getPortTypesByStyle(definitions, serviceName) so style detection runs even for
service-less WSDLs (e.g., find RPC bindings via bindings/portTypes), and update
downstream handling so Message::getPart() is only used for document-style
messages while RPC-style messages are handled with their specific part/parameter
extraction logic (adjust code paths in getPortTypesByStyle and any callers that
assume document-only message parsing).
In
`@core/src/main/java/com/predic8/membrane/core/interceptor/schemavalidation/WSDLSchemaExtractor.java`:
- Around line 63-74: The extracted Document from extractSchema lacks an explicit
base URI; update extractSchema(Element originalSchema, List<Attr>
definitionNamespaces) to accept a base URI parameter (e.g. String location) and
call schema.setDocumentURI(location) before returning (retain
addMissingNamespaceDeclarations(copiedSchema,...)). Also update all callers of
extractSchema to pass the WSDL/schema source location (the same value currently
passed to DOMSource.setSystemId in AbstractXMLSchemaValidator.createValidator())
so xsd:import/include resolution no longer relies solely on downstream
DOMSource.setSystemId.
In
`@core/src/main/java/com/predic8/membrane/core/interceptor/schemavalidation/WSDLValidator.java`:
- Around line 106-129: The response-element whitelist is being applied to SOAP
faults (so declared WSDL faults fail); guard the response-element check to skip
it for faults by changing the condition in WSDLValidator to only run when the
message is a Response AND the validation result is not a fault: i.e. update the
branch that uses message instanceof Response /
isPossibleResponseElement(result.soapElement()) to also require
!result.isFault() (or otherwise short-circuit when result.isFault()), leaving
request-element checks and schema validation unchanged; reference symbols:
skipFaults, result.isFault(), message instanceof Response,
isPossibleResponseElement(...), responseElements.
In
`@core/src/main/java/com/predic8/membrane/core/interceptor/schemavalidation/XMLSchemaValidator.java`:
- Around line 52-54: The getSchemas() method in XMLSchemaValidator currently
returns null which causes a NPE when AbstractXMLSchemaValidator.init() →
createValidators() calls .size(); change getSchemas() to return an empty List
(e.g., Collections.emptyList() or new ArrayList<>()) instead of null so callers
can safely call size() and iterate; update the XMLSchemaValidator.getSchemas()
implementation accordingly.
In
`@core/src/main/java/com/predic8/membrane/core/interceptor/soap/WebServiceExplorerInterceptor.java`:
- Around line 166-175: The operations table code in
WebServiceExplorerInterceptor that maps om -> om.getMessage().getPart() assumes
document-style messages and calls p.getElementQName().toString(), which throws
for RPC messages (elementQName may be null and messages can have multiple
parts); fix by iterating the actual parts collection (use
om.getMessage().getParts() or equivalent) instead of getPart(), for each Part p
build a safe representation that checks p.getElementQName() and falls back to
p.getTypeQName() and p.getName() (avoid calling toString() on a possibly null
QName), and write that representation to the table so RPC multi-part/type-based
signatures are preserved (update the input/output/fault loops that reference
Part p and replace the map/getPart usage accordingly).
- Around line 116-120: The code currently always picks the first WSDL
service/port via service.getPorts().getFirst() which breaks multi-service/port
WSDLs and can throw if that service has no ports; change the selection to locate
the correct service/port instead of hardcoding the first one: when you parse
Definitions.parse(...) iterate over definitions.getServices() to find a service
that matches the proxy's configured service name/namespace or at least has
non-empty ports, then from that service pick a port that best matches the proxy
endpoint (e.g., by port name, binding/endpoint address or other proxy selection
criteria); if no exact match exists, fall back to the first service that
contains ports and the first non-empty port, preserving the later empty-port
fallback logic. Ensure you replace uses of getFirst() (both for service and
port) and update variables service, port and ports accordingly.
In `@core/src/main/java/com/predic8/membrane/core/proxies/SOAPProxy.java`:
- Around line 181-185: getLocation currently always uses
service.getPorts().getFirst() and ignores the proxy's portName and doesn't
handle zero ports; update
getLocation(com.predic8.membrane.core.util.wsdl.parser.Service) to first check
if service.getPorts() is empty and throw a ConfigurationException with a clear
message, then if a portName (the proxy's configured portName) is set, find the
port with that name (e.g., iterate service.getPorts() to match port.getName())
and use its address location, otherwise fall back to the first port; replace the
direct getFirst() call with this selection logic and throw the existing
ConfigurationException if the chosen port has no `@location`.
In `@core/src/main/java/com/predic8/membrane/core/util/SOAPUtil.java`:
- Line 137: The current debug call log.debug("Analyzing SOAP message: {}", msg)
logs the full Message (including headers/body) and may leak sensitive SOAP
payloads; change the call to a static message such as log.debug("Analyzing SOAP
message") and remove passing the Message object (reference: the log.debug
invocation in SOAPUtil.java and the Message instance named msg) so no
Message.toString() is emitted to logs.
In `@core/src/main/java/com/predic8/membrane/core/util/URIUtil.java`:
- Around line 153-176: getNormalizedAbsolutePathOrUri currently treats inputs
with '?' '#' or starting with '//' as URIs and calls URI.create(location) which
can throw or return a relative URI; change this branch to first try creating a
URI and if creation fails or uri.isAbsolute() is false, fall back to treating it
as a filesystem path by calling normalizeInternal(location); if URI creation
succeeds and uri.isAbsolute() then return uri.normalize().toString(); ensure you
reference the existing normalizeInternal(...) method and the URI_SCHEME_PATTERN
logic so the code first attempts URI parsing but gracefully catches
IllegalArgumentException or non-absolute URIs and returns
normalizeInternal(location) instead.
In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Binding.java`:
- Around line 66-69: The code in Binding.getPortType() only compares the local
name from getAttribute("type") which breaks when different namespaces use the
same local name and also will NPE if the attribute is missing; change this to
(1) validate that getAttribute("type") is present and throw a
WSDLParserException with a clear message if it is null/empty, (2) parse the
attribute into a full QName by resolving its prefix to a namespace URI (e.g.
split "prefix:local" or "local" and use the binding/context namespace resolution
helper to get the namespace URI) and create a javax.xml.namespace.QName (or
equivalent) with namespace URI + local part, and (3) match against port types by
comparing both namespaceURI and localPart (e.g.
pt.getQName().equals(parsedQName) or
pt.getName().equals(parsedQName.getLocalPart()) &&
pt.getNamespaceURI().equals(parsedQName.getNamespaceURI())) instead of using
getLocalName(getAttribute("type")); keep use of WSDLParserException for
missing/no-match cases.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Definitions.java`:
- Around line 65-68: The parse() method currently calls getTypes()
unconditionally which makes WSDLs without a <types> block fail early; change
parse() to be tolerant by checking whether getTypes() (or getTypes().element) is
present before calling instantiateXSDElements and importEmbeddedSchemas: only
call instantiateXSDElements(getTypes().element, "schema", Schema.class) and
importEmbeddedSchemas() when getTypes() is non-null (or its element is
non-null), otherwise leave schemas empty/null and continue parsing
targetNamespace, bindings and services; update any code that assumes schemas was
initialized to handle the absent case.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Operation.java`:
- Around line 60-62: The current getMessage() in Operation uses only the local
name (via WSDLParserUtil.getLocalName) which can mis-resolve imported messages
and silently ignore undeclared prefixes; change it to resolve the full QName
from the raw attribute string returned by getAttribute("message"), validate that
any prefix is declared (fail with WSDLParserException if not), extract namespace
URI and local part, and call the definitions-level lookup that accepts
namespace+localName (or otherwise locate by QName) instead of
ctx.definitions().findMessage(localName); update Operation.getMessage() to
perform this QName resolution and throw WSDLParserException on
missing/undeclared prefix or when the fully-qualified message is not found.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/WSDLParserContext.java`:
- Around line 28-30: WSDLParserContext.basePath currently mutates the shared
visitedLocations list before creating the new WSDLParserContext, which can
retroactively change older contexts and break immutable lists; fix by creating a
defensive copy of visitedLocations (e.g., new ArrayList<>(visitedLocations)),
add the new basePath to that copy, and pass the copy into the WSDLParserContext
constructor so each derived context has its own traversal state without mutating
the original list.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/WSDLParserUtil.java`:
- Around line 25-28: The DOM parser in WSDLParserUtil.parse still allows DOCTYPE
and external entity resolution; update the DocumentBuilderFactory and
DocumentBuilder configuration to harden against XXE: enable secure processing
(XMLConstants.FEATURE_SECURE_PROCESSING) on the factory, disable DOCTYPE and
external entity access (features/properties for
"http://apache.org/xml/features/disallow-doctype-decl",
"http://xml.org/sax/features/external-general-entities",
"http://xml.org/sax/features/external-parameter-entities", and set
"http://javax.xml.XMLConstants/property/accessExternalDTD" and
"http://javax.xml.XMLConstants/property/accessExternalSchema" to ""), and ensure
the created DocumentBuilder does not expand external entities (e.g.,
setExpandEntityReferences(false) or equivalent) before calling parse(is); apply
these changes inside the parse method around
DocumentBuilderFactory.newInstance(), dbf.setNamespaceAware(true), and
newDocumentBuilder().parse(is).
In `@core/src/test/java/com/predic8/membrane/core/util/URIUtilTest.java`:
- Around line 241-244: The test method normalizesFileUri should expect the
RFC-compliant URI with three slashes; update the assertion in normalizesFileUri
to expect "file:///test.xml" instead of "file:/test.xml" so
getNormalizedAbsolutePathOrUri(...) matches Java's URI.normalize().toString()
behavior.
In
`@core/src/test/java/com/predic8/membrane/core/ws/relocator/RelocatorTest.java`:
- Around line 41-45: The test methods (e.g., wsdl()) currently only call
relocator.relocate(...) and print os.toString(UTF_8), so they don't assert
behavior; restore meaningful assertions by checking the relocated output and
state: use Relocator.relocate(getFile("/blz-service.wsdl")) and then assert on
relocator's results (e.g., call relocator.isWsdlFound() if still available or
otherwise inspect the relocated output buffer/string via the os.toString(UTF_8)
result) to verify expected content (presence of relocated URLs, expected WSDL
elements, or counts); remove the System.out.println noise and replace it with
assertTrue/assertEquals/assertThat checks that validate the relocation behavior
(apply the same pattern to the other test method that used isWsdlFound()).
In
`@core/src/test/java/com/predic8/membrane/core/ws/relocator/RelocatorWADLTest.java`:
- Around line 1-91: Uncomment the RelocatorWADLTest class and its methods and
either restore it as a runnable test or explicitly disable it: add JUnit 5
annotations (`@BeforeEach` on setUp, `@Test` on testWADLRelocate) and either remove
comments so the test runs, or annotate the class or testWADLRelocate with
`@Disabled`("explain why disabled") (import org.junit.jupiter.api.Disabled).
Ensure the NamespaceContext, Relocator, writer, xpath setup in setUp remains
intact and the helper assertAttribute is visible; keep references to Relocator,
testWADLRelocate, setUp, and assertAttribute so the test compiles and the intent
(or explicit disable reason) is clear.
In `@core/src/test/resources/ws/include/multiple.wsdl`:
- Around line 23-31: The nested xs:element "test" currently sits directly under
xs:element "testResponse" which is invalid; modify the element declaration for
"testResponse" so that its child "test" is defined inside an xs:complexType
containing an xs:sequence (i.e., move the xs:element name="test" into an
xs:complexType/xs:sequence block) ensuring the inner element is wrapped properly
and retains its existing content (the xs:complexType with xs:sequence and
xs:element ref="tns:a").
---
Outside diff comments:
In `@annot/pom.xml`:
- Around line 48-63: The Log4j dependencies log4j-slf4j2-impl and log4j-core in
the annot module are missing test scope: update the dependency entries for
artifactId "log4j-slf4j2-impl" and "log4j-core" to include <scope>test</scope>
so they are only used for tests in the annot module (leave other dependencies
like jackson-databind and jackson-annotations unchanged); ensure you modify the
dependency declarations for those two artifactIds in the pom so they are
test-scoped.
In `@core/src/main/java/com/predic8/membrane/core/ws/relocator/Relocator.java`:
- Around line 43-45: The public constructor currently narrows the API by taking
OutputStreamWriter; revert it to accept java.io.Writer instead (change
Relocator(OutputStreamWriter osw, String protocol, String host, int port, String
contextPath, PathRewriter pathRewriter) to Relocator(Writer writer, ...)), and
pass that Writer into
XMLOutputFactory.newInstance().createXMLEventWriter(writer) so StringWriter and
other Writer implementations continue to work (ensure signature visibility
remains public and update any direct references/tests such as
RelocatorWADLTest).
---
Minor comments:
In
`@core/src/main/java/com/predic8/membrane/core/interceptor/server/WSDLPublisherInterceptor.java`:
- Around line 178-184: getWSDLFromEmbeddingSOAPProxy() currently calls
soapProxy.getWsdl() and unconditionally passes the result to setWsdl(), which
can clear paths and add null into documentsToProcess; add a defensive
null-check: after obtaining wsdl = soapProxy.getWsdl(), if wsdl is null throw a
ConfigurationException (or log and return) with a clear message that the
embedding <soapProxy> has no WSDL, otherwise call setWsdl(wsdl). Update the
method to reference soapProxy and getWsdl() and ensure setWsdl() is only called
with a non-null wsdl to avoid corrupting documentsToProcess.
In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Part.java`:
- Around line 28-37: The public accessors getTypeQName and getElementQName
currently declare non-nullable QName returns but delegate to the `@Nullable`
getAttributeQName; update their signatures to return `@Nullable` QName (i.e.
annotate both getTypeQName and getElementQName with the same `@Nullable` used on
getAttributeQName) so the public API matches implementation, and add any
required import for the chosen `@Nullable` annotation if missing.
In `@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/Port.java`:
- Around line 27-29: The getAddress() method currently calls
instantiateElements(element, "address", Address.class).getFirst() which can
throw NoSuchElementException when no <address> child exists; update getAddress()
to defensively handle the empty case like getBinding() does by checking the
instantiated list for emptiness and returning null (or an appropriate absent
value) instead of calling getFirst() directly so missing address elements don't
cause unchecked exceptions.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/ProtocolOperation.java`:
- Around line 1-16: The file ProtocolOperation.java is missing the Apache 2.0
license header; add the same standard Apache 2.0 header used in other new files
(e.g., Address.java, OperationTest.java) at the top of the file before the
package declaration so the class ProtocolOperation and its constructor/methods
(ProtocolOperation(WSDLParserContext, Node) and getSoapAction()) have the
required license comment header.
In `@core/src/test/resources/ws/import/message-references-import.wsdl`:
- Around line 37-38: The binding declares an operation named "test" but the
portType defines the operation as "get", causing a name mismatch; update the
binding <operation name="test"> to use the portType operation name "get" (or
alternatively rename the portType operation to "test" if that was intended) and
ensure the corresponding soap:operation soapAction is updated to match the
chosen operation name so the binding correctly references the portType
operation.
- Around line 23-25: The WSDL message GetResponse references a non-existent
element m:Response; update the part element reference in the GetResponse message
to the correct element defined in messages.xsd (m:getResponse) so the part
attribute uses the existing element name (modify the <message
name="GetResponse"> part element reference from m:Response to m:getResponse).
---
Nitpick comments:
In
`@core/src/main/java/com/predic8/membrane/core/interceptor/schemavalidation/ValidatorInterceptor.java`:
- Around line 153-166: The empty log.error calls in the ReadingBodyException
catch and the generic Exception catch around validator.validateMessage remove
context; change them to include a short static prefix (e.g.
"ValidatorInterceptor: could not read message body" in the ReadingBodyException
block and "ValidatorInterceptor: validation failed" in the
validator.validateMessage catch) while still passing the exception object to
log.error; update the log.error invocations near the ReadingBodyException
handler and the catch(Exception e) after validator.validateMessage() to use
these descriptive messages so the internal(...).detail(...).exception(e) flow
remains unchanged.
In `@core/src/main/java/com/predic8/membrane/core/util/soap/WSDLUtil.java`:
- Around line 27-30: rewriteRelativeWsdlPath currently calls matcher(path) and
formatted(replacementName) without guarding against nulls; add null-safety at
the start of the public method by validating inputs (e.g., use
Objects.requireNonNull(path, "path must not be null") and either
requireNonNull(replacementName, "replacementName must not be null") or coerce
replacementName to "" when null) so matcher(...) never receives null and
formatted(...) does not produce the literal "null"; keep the rest of the method
(relativePathPattern.matcher(...).replaceAll(quoteReplacement(...))) unchanged
and throw clear IllegalArgumentException/NullPointerException messages if you
choose to reject nulls.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/BindingStyle.java`:
- Around line 1-6: Add the standard Apache 2.0 license header to the top of
BindingStyle.java (the file in package
com.predic8.membrane.core.util.wsdl.parser) so it matches other new files in the
PR; insert the same multi-line Apache 2.0 copyright/license block used elsewhere
in the repo above the package declaration in BindingStyle.java.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/schema/AbstractIncludeImport.java`:
- Around line 49-51: The catch block in AbstractIncludeImport currently rethrows
a generic RuntimeException losing context; update the catch to throw a new
RuntimeException that includes the schemaLocation (and any identifying info) in
the message and pass the original exception as the cause (e.g., throw new
RuntimeException("Failed to resolve schemaLocation: "+ schemaLocation, e)); this
keeps the original stacktrace and makes it clear which schema failed to resolve.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/schema/Import.java`:
- Around line 52-62: The null/empty check for schemaLocation is duplicated;
update getSchema(WSDLParserContext ctx) to call the existing hasSchemaLocation()
instead of repeating the condition: if hasSchemaLocation() returns false, return
null, otherwise delegate to super.getSchema(ctx). Keep the existing
hasSchemaLocation() method as the single source of truth for the empty-check so
future changes only need to be made in one place.
In
`@core/src/main/java/com/predic8/membrane/core/util/wsdl/parser/WSDLParserException.java`:
- Around line 17-20: WSDLParserException currently only has a String constructor
so underlying parse exceptions lose stack/context; add overloaded constructors
to preserve causes: implement WSDLParserException(String message, Throwable
cause) that calls super(message, cause) and optionally
WSDLParserException(Throwable cause) that calls super(cause) (and keep the
existing WSDLParserException(String) for compatibility) so callers can
wrap/paraphrase parse failures without losing the original exception.
In
`@core/src/test/java/com/predic8/membrane/core/interceptor/schemavalidation/SOAPFaultTest.java`:
- Line 39: Replace the explicit local variable type with "var" in the test
methods that still declare a ValidatorInterceptor explicitly (look for local
variables named like the interceptor in testSkipFaults and testSkipFault2) so
that they match the style used when assigning createValidatorInterceptor(false);
i.e., change "ValidatorInterceptor i = ..." to "var i = ..." in those test
methods.
In
`@core/src/test/java/com/predic8/membrane/core/interceptor/schemavalidation/SOAPUtilTest.java`:
- Around line 127-129: The test's getMessage method currently opens files via
new FileInputStream(path) which makes tests reliant on the working directory;
change getMessage to load the test file from the classpath (e.g., use
getClass().getResourceAsStream(path) or
Thread.currentThread().getContextClassLoader().getResourceAsStream(path)) and
pass that InputStream into ok().contentType(TEXT_XML).body(...).build(); then
update all call sites that pass filesystem-relative strings to use
classpath-relative resource paths such as "wsdlValidator/soapFaultFromSpec.xml"
and "wsdlValidator/soapFaultCustom.xml" so resources are resolved consistently
in IDE/Maven/CI.
In `@core/src/test/java/com/predic8/membrane/core/ws/WSDLIncludeImportTest.java`:
- Around line 102-104: Wrap the test resource lookup in an explicit null-check
using Objects.requireNonNull to avoid an ambiguous NPE: call
Objects.requireNonNull(WSDLIncludeImportTest.class.getResource("/ws/include/include.wsdl"),
"resource '/ws/include/include.wsdl' not found") and then call toString() for
the WSDLValidator constructor; also add the java.util.Objects import if missing.
This change targets the WSDLIncludeImportTest test where WSDLValidator(...) is
constructed.
In `@core/src/test/resources/ws/include/include.xsd`:
- Around line 6-16: Remove the commented-out schema directives to clean up the
fixture: delete the commented lines containing "<!-- <xs:include
schemaLocation="./xsd/messages.xsd"/> -->" and "<!-- <xs:import
namespace="http://example.com/test/messages"
schemaLocation="./xsd/messages.xsd"/> -->" so only the active xs:include
(schemaLocation="xsd/inc/MessageStructureTypes2.xsd") and the xs:element
definitions "test" and "testResponse" remain, improving readability and
maintainability.
In `@core/src/test/resources/ws/include/xsd/inc/MessageStructureTypes2.xsd`:
- Around line 1-8: The XSD imports the data-types namespace but never uses it;
update MessageStructureTypes2.xsd so the imported namespace is referenced by
declaring a prefix (e.g., xmlns:dt="http://example.com/test/data-types") and
change the xs:element named "du" to reference a type or element from that
namespace (for example set du to type="dt:SomeType" or include a child element
from the dt: namespace) so cross-namespace resolution is exercised; locate the
xs:element name="du" and the xs:import statement and add the xmlns:dt
declaration and the dt: reference there.
In `@pom.xml`:
- Around line 273-287: The POM has mixed XMLUnit versions: xmlunit-core is
2.10.0 while xmlunit-matchers remains 2.9.1; update the xmlunit-matchers
dependency entry (artifactId: xmlunit-matchers) to use version 2.10.0 so both
modules align with xmlunit-core, then run the build/tests to verify
compatibility.
Contributor
✅ Actions performedReviews resumed. |
…d simplify test setup - Replaced `ServiceProxy` with `APIProxy` for target configuration. - Improved test readability by using string-based SOAP request in place of external XML files. - Streamlined setup and teardown with updated Router lifecycle methods.
- Enhanced base path handling in `WSDLParserContext` to avoid alterations in original visited locations. - Simplified schema parsing logic in `Definitions` with conditional `instantiateChild`. - Updated `WSDLValidator` to correctly handle fault cases in response element validation. - Added WSDL namespace detection in `Relocator` and respective tests in `RelocatorTest`. - Restored and partially reintroduced `RelocatorWADLTest` with Apache License header.
rrayst
approved these changes
Mar 15, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary by CodeRabbit
New Features
Bug Fixes
Dependencies