28857 openapi include exclude #29769

benjamin-confino · 2024-10-02T15:22:21Z

I have considered the risk of behavior change or other zero migration impact (https://github.com/OpenLiberty/open-liberty/wiki/Behavior-Changes).
If this PR fixes an Issue, the description includes "Fixes #FILLMEIN" or "Resolves #FILLMEIN" (verify release bug label if applicable: https://github.com/OpenLiberty/open-liberty/wiki/Open-Liberty-Conventions).
If this PR resolves an external Known Issue (including APARS), the description includes "Fixes #FILLMEIN" or "Resolves #FILLMEIN".

resolves #28857

This PR

Creates a centralized service for openAPI to read the relevent parts of the server.xml
Uses that service to read what applications/modules are included or excluded, overriding mpConfig if both are in use.

LibbyBot · 2024-10-02T18:59:14Z

Code analysis and actions

DO NOT DELETE THIS COMMENT.

11 product code files were changed.
Please describe in a separate comment how you tested your changes.
1 test infrastructure code files were changed.
Test failures/errors in the build could be due to these changes.
12 FAT files were changed, added, or removed.
Check that the build did not break the affected FAT suite(s).
1 NLS files were changed and need an ID review.
@OpenLiberty/message-reviewer Please review.
dev/io.openliberty.microprofile.openapi.internal.common/resources/OSGI-INF/l10n/metatype.properties

jantley-ibm

ID review comments attached; there appears to be a placeholder reminder in a couple of spots, so when that is completed I will review that text.
If you have questions or concerns about any of the suggested changes, please let me know.
I will add the "ID reviewed" label when we agree on changes. Thanks!

jantley-ibm · 2024-10-03T13:18:25Z

...openliberty.microprofile.openapi.internal.common/resources/OSGI-INF/l10n/metatype.properties

+
+include=MicroProfile OpenAPI included modules.
+include=A comma separated list of modules that will be included in the OpenAPI document. Modules are in the format of {ApplicationName}/{ModuleName} or {ApplicationName} which refer to to <TODO, sort out the confusion of names>. An ApplicationName which does not specify a module will include all modules in the application. The special value "all" will include all available modules.


Following appears to be a note that further work needs to be done:
to to <TODO, sort out the confusion of names>

Same appears in line 32

Change
comma separated
to
comma-separated

Change
that will be included
to
that are to be included

Change
An ApplicationName which
to
An ApplicationName that

Change
will include all modules
to
includes all modules

Change
The special value "all" will include
to
The "all" special value includes

jantley-ibm · 2024-10-03T13:24:41Z

...openliberty.microprofile.openapi.internal.common/resources/OSGI-INF/l10n/metatype.properties

+
+exclude=MicroProfile OpenAPI excluded modules.
+exclude=A comma separated list of that will be excluded from the OpenAPI document. Modules are in the format of {ApplicationName}/{ModuleName} or {ApplicationName} which refer to to <TODO, sort out the confusion of names>. An ApplicationName which does not specify a module will exclude all modules in the application. An exclude takes precedent over an include.


Following appears to be a note that further work needs to be done:
to to <TODO, sort out the confusion of names>

Change
A comma separated list of that will be excluded
to
A comma-separated list of modules that are excluded

Change
An exclude takes precedent over an include.
to
Exclusion supersedes inclusion.

benjamin-confino · 2024-10-09T23:01:58Z

#build

LibbyBot · 2024-10-09T23:02:05Z

Your personal pipeline request is at https://libh-proxy1.fyre.ibm.com/cognitive/pipelineAnalysis.html?uuid=8008f055-9474-4534-83f0-158272300a50

Target locations of links might be accessible only to IBM employees.

LibbyBot · 2024-10-10T08:04:13Z

https://libh-proxy1.fyre.ibm.com/cognitive/pipelineAnalysis.html?uuid=8008f055-9474-4534-83f0-158272300a50 has errors/failures.

benjamin-confino · 2024-10-10T10:33:20Z

#build

LibbyBot · 2024-10-10T10:33:27Z

Your personal pipeline request is at https://libh-proxy1.fyre.ibm.com/cognitive/pipelineAnalysis.html?uuid=b9f42711-27e0-4f75-ba46-0a8d69789982

Target locations of links might be accessible only to IBM employees.

LibbyBot · 2024-10-10T10:36:33Z

Your personal build request is at https://wasrtc.hursley.ibm.com:9443/jazz/resource/itemOid/com.ibm.team.build.BuildResult/_IoGZUIbrEe-vDLerZarrlA

Target locations of links might be accessible only to IBM employees.

...0.internal/src/io/openliberty/microprofile/openapi20/internal/ModuleSelectionConfigImpl.java

...api.2.0.internal/src/io/openliberty/microprofile/openapi20/internal/merge/ModelEquality.java

...napi.2.0.internal/src/io/openliberty/microprofile/openapi20/internal/utils/OpenAPIUtils.java

Azquelt · 2024-10-10T15:15:45Z

dev/com.ibm.ws.cdi.api_fat/autoFVT.zip

Remove this.

Azquelt · 2024-10-10T15:18:50Z

dev/fattest.simplicity/src/componenttest/topology/impl/LibertyServer.java

I would suggest moving this change into a separate PR as it's the only change in fattest.simplicity.

Changing any of the core test framework projects will mean your personal builds need to run every bucket, rather than only the ones which use your feature and therefore take a while longer to complete.

I wouldn't shy away from making changes to the test framework if they're needed for your PR, but if it's just doc changes like this I'd move it to its own PR to avoid the larger PB.

Azquelt · 2024-10-10T16:00:12Z

...c/io/openliberty/microprofile/openapi/internal/common/services/OpenAPIAppConfigProvider.java

+    public static interface OpenAPIAppConfigListener extends Comparable<OpenAPIAppConfigListener> {
+        public void processConfigUpdate();
+
+        public int getConfigListenerPriority();
+    }


This interface needs documented.

If it implements Comparable, it needs to also document how two objects are compared and whether that comparison is consistent with equals. However, in this case I think it would make more sense not to make this implement Comparable, and instead use a Comparator which compares based on the value returned by getConfigListenerPriority().

In addition, please document whether listeners are called with the largest priority first or the smallest priority first.

LibbyBot · 2024-10-10T19:19:36Z

Your personal build request is at https://wasrtc.hursley.ibm.com:9443/jazz/resource/itemOid/com.ibm.team.build.BuildResult/_NFwGMIc0Ee-vDLerZarrlA

Target locations of links might be accessible only to IBM employees.

LibbyBot · 2024-10-10T20:57:56Z

https://libh-proxy1.fyre.ibm.com/cognitive/pipelineAnalysis.html?uuid=b9f42711-27e0-4f75-ba46-0a8d69789982 has errors/failures.

LibbyBot · 2024-10-11T04:33:53Z

The build benjamin-confino-29769-20241010-0344
https://wasrtc.hursley.ibm.com:9443/jazz/resource/itemOid/com.ibm.team.build.BuildResult/_IoGZUIbrEe-vDLerZarrlA
completed and has errors or failures.

For help analyzing your personal build, go to https://libh-proxy1.fyre.ibm.com/cognitive/buildAnalysis.html?uuid=_IoGZUIbrEe-vDLerZarrlA

Azquelt · 2024-10-11T10:18:06Z

dev/io.openliberty.microprofile.openapi.2.0.internal/bnd.bnd

-# Copyright (c) 2020, 2024 IBM Corporation and others.
+# Copyright (c) 2020, 2022 IBM Corporation and others.


Incorrect copyright date change?

Azquelt · 2024-10-11T10:18:45Z

...2.0.internal/src/io/openliberty/microprofile/openapi20/internal/ApplicationRegistryImpl.java

+    @SuppressWarnings("unused")
+    private OpenAPIAppConfigProvider openAPIAppConfigProvider;


Remove SuppressWarnings

Azquelt · 2024-10-11T10:19:13Z

...2.0.internal/src/io/openliberty/microprofile/openapi20/internal/ApplicationRegistryImpl.java

+    @SuppressWarnings("unused")
+    private OpenAPIAppConfigProvider openAPIAppConfigProvider;
+
+    @Reference(cardinality = ReferenceCardinality.MANDATORY, policy = ReferencePolicy.STATIC, unbind = "unbindAppConfigListener")


unbind is not necessary here, it gets inferred based on the method names.

Azquelt · 2024-10-11T10:20:33Z

...2.0.internal/src/io/openliberty/microprofile/openapi20/internal/ApplicationRegistryImpl.java

+    public void bindAppConfigListener(OpenAPIAppConfigProvider openAPIAppConfigProvider) {
+        this.openAPIAppConfigProvider = openAPIAppConfigProvider;
+        openAPIAppConfigProvider.registerAppConfigListener(this);
+    }
+
+    public void unbindAppConfigListener(OpenAPIAppConfigProvider openAPIAppConfigProvider) {
+        openAPIAppConfigProvider.unregisterAppConfigListener(this);
+        this.openAPIAppConfigProvider = null;
+    }
+
+    @Reference
+    public void bindOpenAPIAppConfigProvider(OpenAPIAppConfigProvider openAPIAppConfigProvider) {
+        this.openAPIAppConfigProvider = null;
+        openAPIAppConfigProvider.registerAppConfigListener(this);
+    }


This looks like you have two references to OpenAPIAppConfigProvider, but the first one is named appConfigListener.

Azquelt · 2024-10-11T10:41:19Z

...0.internal/src/io/openliberty/microprofile/openapi20/internal/ModuleSelectionConfigImpl.java

+    @Reference(cardinality = ReferenceCardinality.MANDATORY, policy = ReferencePolicy.STATIC, unbind = "unbindAppConfigListener")
+    public void bindAppConfigListener(OpenAPIAppConfigProvider openAPIAppConfigProvider) {
+        this.configFromServerXML = openAPIAppConfigProvider;
+        openAPIAppConfigProvider.registerAppConfigListener(this);
+    }
+
+    public void unbindAppConfigListener(OpenAPIAppConfigProvider openAPIAppConfigProvider) {
+        openAPIAppConfigProvider.unregisterAppConfigListener(this);
+        this.configFromServerXML = null;
+    }


Since this is a static reference, it might be easier to annotate the field and register as a listener in activate/deactivate.

Technically here we're registering as a listener before we're activated.

Azquelt · 2024-10-11T10:54:13Z

...0.internal/src/io/openliberty/microprofile/openapi20/internal/ModuleSelectionConfigImpl.java

+    @Override
+    public synchronized void processConfigUpdate() {
+        //Reset everything for the next lazyInit
+        hasBeenInit = false;
+        isAll = false;
+        isFirst = false;
+        serverxmlmode = false;
+        included = null;
+        excluded = null;
+    }


We have an issue with thread safety because we have this method to reset everything and re-initialize on next use.

If lazyInit was only called once it would be sufficient for only the lazyInit method to be synchronized as long as every method which needs to access any of these values calls lazyInit before doing so.

However, since we have the ability to reset, every use of any of these variables must be synchronized, otherwise they could be reset in the middle of a method which uses them.

To avoid this, I would suggest:

add a new static inner class to hold these values - let's call it ConfigValues

rather than having a lazyInit method which sets all these fields, have a method which creates, caches and returns a ConfigValues object

all other code which needs the ConfigValues object must retrieve it via this method

processConfigUpdate then only needs to delete the cached ConfigValues object

this method and the method to get ConfigValues may need synchronization or correct use of volatile, but everything else should be correct without synchronization

Azquelt · 2024-10-11T10:55:49Z

...napi.2.0.internal/src/io/openliberty/microprofile/openapi20/internal/utils/OpenAPIUtils.java

+
+        System.out.println("GREPGREP " + a + " " + b + " " + comparator.getClass().toGenericString());
+


Remove this (or make it proper entry trace)

Azquelt · 2024-10-11T10:58:57Z

...0.internal/src/io/openliberty/microprofile/openapi20/internal/ModuleSelectionConfigImpl.java

    @Override
    public String toString() {
+        lazyInit();
        StringBuilder sb = new StringBuilder("Module Selection Config[");
        if (isFirst) {


I'm a bit concerned about toString() doing initialization here. That's the sort of thing that can cause enabling trace to change behaviour...

Could we make it so that toString() reports "Uninitialized Config` or something similar if initialization hasn't happened yet?

Azquelt · 2024-10-11T11:01:33Z

...0.internal/src/io/openliberty/microprofile/openapi20/internal/ModuleSelectionConfigImpl.java

+        if (serverxmlmode && ProductInfo.getBetaEdition()) {
+            for (Iterator<ModuleName> iterator = includedNotYetSeen.iterator(); iterator.hasNext();) {
+                ModuleName moduleName = iterator.next();
+                for (ModuleInfo moduleInfo : moduleInfos) {
+                    if (matches(moduleName, moduleInfo, false)) {
+                        includedNotYetSeenButSeenUnderOldNaming.add(moduleName);
+                        iterator.remove();
+                        break;
+                    }
+                }
+            }
+
+            for (String unmatchedInclude : includedNotYetSeenButSeenUnderOldNaming.stream().map(ModuleName::toString).collect(toList())) {
+                Tr.warning(tc, MessageConstants.OPENAPI_USING_WRONG_NAME_SOURCE_CWWKO1680W, Constants.MERGE_INCLUDE_CONFIG, unmatchedInclude);
+            }
+        }


Can we do this check that the may have used the wrong name for excluded apps as well as included apps please?

(The other check for a name that's missing should remain as only checking the included apps).

....2.0.internal/src/io/openliberty/microprofile/openapi20/internal/utils/MessageConstants.java

Co-authored-by: Andrew Rouse <anrouse@uk.ibm.com>

Azquelt · 2024-10-11T11:07:53Z

...2.0.internal/src/io/openliberty/microprofile/openapi20/internal/ApplicationRegistryImpl.java

+        Map<String, ApplicationRecord> oldApps = applications;
+        applications = new LinkedHashMap<>();
+        for (ApplicationRecord record : oldApps.values()) {
+            //Add application uses config to decide if it creates and registers any providers in ApplicationInfo
+            //Rather than map from the old state to the new state when the config changes, KISS and start again.
+            addApplication(record.info);
+        }
+        cachedProvider = null;


In general, we use synchronized (this) in the method body rather than synchronized in the signature so that the generated trace is outside the synchronized block, making it easier to debug lock contention.

Since addApplication can take a while to complete, I think we should do that here.

Azquelt · 2024-10-11T11:08:50Z

...2.0.internal/src/io/openliberty/microprofile/openapi20/internal/ApplicationRegistryImpl.java

+    public int getConfigListenerPriority() {
+        return 2;
+    }


Comment that this is to ensure we're called after the ModuleSelectionConfig

Azquelt · 2024-10-11T11:13:01Z

...on/src/io/openliberty/microprofile/openapi/internal/common/OpenAPIAppConfigProviderImpl.java

+    @Modified
+    protected void modified(Map<String, Object> properties) {
+        if (ProductInfo.getBetaEdition()) {
+            if (TraceComponent.isAnyTracingEnabled() && tc.isEventEnabled()) {
+                Tr.event(tc, "Processing update to server.xml");
+            }
+            processModuleConfig(properties);
+            for (OpenAPIAppConfigListener listener : openAPIAppConfigListeners) {
+                listener.processConfigUpdate();
+            }
+        }
+    }


This class only cares about a subset of properties from the <mpOpenAPI> config element. If someone changes a config value that isn't related to module inclusion/exclusion then we shouldn't call the listeners since re-generating all the OpenAPI documents is a relatively expensive operation.

Azquelt · 2024-10-11T12:40:18Z

...2.0.internal_fat/fat/src/io/openliberty/microprofile/openapi20/fat/deployments/test1/web.xml

+<?xml version="1.0" encoding="UTF-8"?>
+<!-- Used for tests that need a non-default deployment name for a war -->
+<web-app version="3.0"
+        xmlns="http://java.sun.com/xml/ns/javaee"
+        xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+        xsi:schemaLocation="http://java.sun.com/xml/ns/javaee web-app_3_0.xsd"
+        id="testID">
+        <module-name>nameFromWar</module-name>
+</web-app>


Please use WebXmlAsset instead of having a web.xml file included manually.

Azquelt · 2024-10-11T12:44:46Z

.../src/io/openliberty/microprofile/openapi20/fat/deployments/MergeServerXMLTestWithUpdate.java

+        server.swapInServerXMLFromPublish(SERVER_XML_NAME);
+        server.waitForStringInLogUsingMark("CWWKG0017I"); //The server configuration was successfully updated


Suggested change

server.swapInServerXMLFromPublish(SERVER_XML_NAME);

server.waitForStringInLogUsingMark("CWWKG0017I"); //The server configuration was successfully updated

server.setMarkToEndOfLog();

server.swapInServerXMLFromPublish(SERVER_XML_NAME);

server.waitForConfigUpdateInLogUsingMark(null);

I think we need to manually set the mark ourselves before starting the update if we want to use it afterwards.

Also there's a specific method for waiting for config updates.

But also, I would prefer it if we had a method to update the config that made it more obvious in the test what the value of the config is so you don't have to jump between lots of files to understand the test.

E.g. something that let you make calls like this:

updateConfig(server, includeApp("all")); updateConfig(server, includeApp("app1")); updateConfig(server, includeApp("app1"), includeApp("app2"), excludeModule("app2/war1"));

or like this:

openapiConfig(server).includeApp("all").apply(); openapiConfig(server).includeApp("app1").apply(); openapiConfig(server).includeApp("app1") .includeApp("app2") .excludeModule("app2/war1") .apply();

Edit: I've since commented on MergeServerXmlTest that we should probably cover the same cases that are covered in MergeConfigTest. Those tests could be conducted using dynamic updates, though I'm unsure whether it would cover the same cases you have in this class where there are apps deployed and you update the config.

I wanted to update this message though in case you were working through and implemented this suggestion before seeing the other comment.

Azquelt · 2024-10-11T12:45:26Z

.../src/io/openliberty/microprofile/openapi20/fat/deployments/MergeServerXMLTestWithUpdate.java

+        server.swapInServerXMLFromPublish(APP1_XML_NAME);
+        server.waitForStringInLogUsingMark("CWWKG0017I"); //The server configuration was successfully updated


Suggested change

server.swapInServerXMLFromPublish(APP1_XML_NAME);

server.waitForStringInLogUsingMark("CWWKG0017I"); //The server configuration was successfully updated

server.setMarkToEndOfLog();

server.swapInServerXMLFromPublish(APP1_XML_NAME);

server.waitForConfigUpdateInLogUsingMark(null);

...rty.microprofile.openapi.2.0.internal_fat/publish/servers/OpenAPIMergeTestServer/jvm.options

Azquelt · 2024-10-11T13:42:37Z

...at/src/io/openliberty/microprofile/openapi20/fat/deployments/StartupWarningMessagesTest.java

+    @Test
+    public void tesWarningWhenMatchesOldNamingScheme() throws Exception {
+
+        EnterpriseArchive ear = null;
+        WebArchive war1 = ShrinkWrap.create(WebArchive.class, "test1.war")
+                                    .addClasses(DeploymentTestApp.class, DeploymentTestResource.class);
+
+        WebArchive war2 = ShrinkWrap.create(WebArchive.class, "test2.war")
+                                    .addClasses(DeploymentTestApp.class, DeploymentTestResource.class);
+
+        WebArchive war3 = ShrinkWrap.create(WebArchive.class, "test3.war")
+                                    .addClasses(DeploymentTestApp.class, DeploymentTestResource.class)
+                                    .addAsWebInfResource(DeploymentTestApp.class.getPackage(), "web.xml", "web.xml");
+
+        ear = ShrinkWrap.create(EnterpriseArchive.class, "testEar.ear")
+                        .addAsModules(war1, war2, war3);
+        ShrinkHelper.exportAppToServer(server, ear, SERVER_ONLY, DISABLE_VALIDATION);
+
+        server.setAdditionalSystemProperties(
+                                             Collections.singletonMap("mp_openapi_extensions_liberty_merged_include", "testEar"));
+        server.startServer();
+
+        //Test that a warning message saying we cant find the app but could under the pre OpenAPI 4.0 rules 
+        assertThat(
+                   server.findStringsInLogs(
+                                            "CWWKO1680W"),
+                   hasSize(1));
+
+    }


This test method contains lots of things are aren't relevant and hides or doesn't explain everything important about the test.

I don't think it's relevant that the app is an ear, or that it contains three modules, or that one of the modules contains a web.xml. Please simplify this.

I'm unsure if the mp_openapi_extensions_liberty_merged_include property is important - this looks to be identical to the config in the server.xml? Should it be removed?

The server.xml gives a custom name to the application and adds configuration which includes the app using the wrong name. These details should either be explained in a comment or be moved into the test method.

The error message appears because the server.xml name is "nameFromServerXml" while the deployment descriptor name is "testEar" because there's no deployment descriptor so it defaults to the archive name without the extension. This detail should be explained in a comment.

There's no description of what the CWWKO1680W message is or why we expect it.

Azquelt · 2024-10-11T13:46:35Z

...at/src/io/openliberty/microprofile/openapi20/fat/deployments/StartupWarningMessagesTest.java

+        //Test that a warning message saying we cant find the app but could under the pre OpenAPI 4.0 rules 
+        assertThat(
+                   server.findStringsInLogs(
+                                            "CWWKO1680W"),
+                   hasSize(1));


The warning message is wrong, the rules differ depending on whether the config came from server.xml or MP Config, not whether we're using mpOpenAPI-4.0 or an earlier feature.

We need to ensure the message has been substituted correctly as the name of the message also includes CWWKO1680W. Since this message has placeholders, we should include the expected values for those placeholders in the regex.

I'm not sure if this message is guaranteed to have been written by the time that server.startServer() returns. We should use server.waitForStringInLogs here unless we're certain that waiting is never necessary.

Azquelt · 2024-10-11T13:49:10Z

...at/src/io/openliberty/microprofile/openapi20/fat/deployments/StartupWarningMessagesTest.java

+        ear = ShrinkWrap.create(EnterpriseArchive.class, "testEar.ear")
+                        .addAsModules(war1, war2, war3);
+        ShrinkHelper.exportAppToServer(server, ear, SERVER_ONLY, DISABLE_VALIDATION);


We're disabling validation here because the name of the app doesn't match the archive name. That's probably worth a comment.

Since we do expect the app to start, we should do the validation manually by calling server.addInstalledAppForValidation("nameFromServerXml"); ourselves after exporting the app.

Azquelt · 2024-10-11T14:29:19Z

...al_fat/fat/src/io/openliberty/microprofile/openapi20/fat/deployments/MergeServerXMLTest.java

+@RunWith(FATRunner.class)
+public class MergeServerXMLTest {


This test class appears to have been adapted from MergeTest?

MergeTest is meant to test that we correctly merge documents, all tests run with a simple config which merges all apps.

I don't think we need to repeat the same tests with with the same basic config in server.xml.

MergeConfigTest is our main test for the merging config - we should include all the include and exclude cases from there in our server.xml config testing. Since the server.xml config is dynamic, we won't have to restart the server between each test like we do in MergeConfigTest.

The new cases that you've covered in this class where you're updating the config are obviously still valuable.

Azquelt · 2024-10-11T14:31:50Z

...pi.2.0.internal/src/io/openliberty/microprofile/openapi20/internal/ApplicationProcessor.java

@@ -48,6 +48,7 @@
 import io.openliberty.microprofile.openapi20.internal.cache.ConfigSerializer;
 import io.openliberty.microprofile.openapi20.internal.services.ConfigFieldProvider;
 import io.openliberty.microprofile.openapi20.internal.services.ModelGenerator;
+import io.openliberty.microprofile.openapi20.internal.services.ModuleSelectionConfig;


spurious import

Azquelt · 2024-10-11T14:52:50Z

...0.internal/src/io/openliberty/microprofile/openapi20/internal/ModuleSelectionConfigImpl.java

+            for (Iterator<ModuleName> iterator = includedNotYetSeen.iterator(); iterator.hasNext();) {
+                ModuleName moduleName = iterator.next();
+                for (ModuleInfo moduleInfo : moduleInfos) {
+                    if (matches(moduleName, moduleInfo, false)) {
+                        includedNotYetSeenButSeenUnderOldNaming.add(moduleName);
+                        iterator.remove();
+                        break;
+                    }
+                }
+            }
+
+            for (String unmatchedInclude : includedNotYetSeenButSeenUnderOldNaming.stream().map(ModuleName::toString).collect(toList())) {
+                Tr.warning(tc, MessageConstants.OPENAPI_USING_WRONG_NAME_SOURCE_CWWKO1680W, Constants.MERGE_INCLUDE_CONFIG, unmatchedInclude);
+            }


This code needs updating because the placeholders in the message have been updated.

benjamin-confino marked this pull request as draft October 2, 2024 15:22

LibbyBot added the CLA Signed label Oct 2, 2024

jantley-ibm reviewed Oct 3, 2024

View reviewed changes

dmuelle added the ID waiting on development label Oct 4, 2024

benjamin-confino force-pushed the 28857-openapi-include-exclude branch 5 times, most recently from 4595c11 to 6af9c5b Compare October 9, 2024 15:59

benjamin-confino force-pushed the 28857-openapi-include-exclude branch 2 times, most recently from 269270e to 5ce8c23 Compare October 10, 2024 08:55

Get config from server.xml

a5b2a72

benjamin-confino force-pushed the 28857-openapi-include-exclude branch 2 times, most recently from 404b48c to e5127e8 Compare October 10, 2024 10:31

benjamin-confino added the in:MicroProfile/OpenAPI label Oct 10, 2024

create fat test for merging with server.xml

199d8a3

benjamin-confino force-pushed the 28857-openapi-include-exclude branch from e5127e8 to 199d8a3 Compare October 10, 2024 11:52