Update SmallRye Config to 3.8.1

[radcortez]

SmallRye Config Release notes:

3.8.0

• #1153 Release 3.8.0
• #1150 Update SR Commons to 2.4.0
• #1149 Consistent collection type for profiles
• #1148 Always include super properties of mappings
• #1147 PropertyName do not match star with empty
• #1146 Generate mapping names
• #1145 Bump org.eclipse.parsson:parsson from 1.1.5 to 1.1.6
• #1144 Bump io.smallrye:smallrye-parent from 42 to 43
• #1143 Support expression expansion with secret keys configuration values
• #1142 Bump org.ow2.asm:asm from 9.6 to 9.7
• #1101 Bug! Spring bootJar: NoSuchFileException: /path/application.jar!/BOOT-INF/classes, /path/application.jar!/BOOT-INF/lib/somelib.jar

3.8.1

• #1159 Release 3.8.1
• #1158 Keep all mapping names
• #1157 Check invalid mapped properties
• #1156 Override super members defaults
• #1155 Bump org.apache.maven.plugins:maven-shade-plugin from 3.5.2 to 3.5.3
• #1154 Change getSuperTypes to public

Quarkus:

• Adjusted API integration to not require @ConfigMapping method reflection registration

[gsmet]

@zakkak @galderz @yrodiere @maxandersen some good news on the reflection front:

org.opentest4j.AssertionFailedError: Expected analysis_results.methods.reflection to be within range [4893 +- 3%] but was 4551 ==> expected: but was:

It fails because we are below the threshold :).

Now, this is just the methods metrics but it was the one growing significantly every time we migrated a config class to @ConfigMapping.
As we don't use Hibernate Validator constraints in the core config, it should avoid registering the methods for reflection for all the core config move to @ConfigMapping.

@radcortez needs to adjust the testing a bit on his side of the PR but we should be good for 3.11.

[zakkak]

That's great, well done!

[yrodiere]

Great!

Just a heads-up: updates to the expected metrics will likely conflict with #40102, so you might want to wait until that's merged before doing all that work.

[gsmet]

@zakkak just to be clear, it's only improving the methods (I think), so https://github.com/quarkusio/quarkus/pull/40102/files#diff-f740c865e615865a3cf3a4018a7115b9e528d6ad48595ec8e264bb4fc5b2ea83R4 is probably still problematic.

[quarkus-bot]

Status for workflow Quarkus CI

This is the status report for running Quarkus CI on commit 0a775e3.

✅ The latest workflow run for the pull request has completed successfully.

It should be safe to merge provided you have a look at the other checks in the summary.

You can consult the Develocity build scans.

[gsmet]

Thanks for collaborating with me on this, @radcortez !
It's definitely something we needed to improve before considering more candidates for @ConfigMapping.

[michalvavrik]

Hello @radcortez , we have experienced slight change of behavior that I describe here quarkus-qe/quarkus-test-framework#1117. I don't think it is a bug, but the way I read it in past environment variable MP_MESSAGING_INCOMING_CHANNEL_STOCK_PRICE_CONNECTOR was translated to mp.messaging.incoming.channel-stock-price.connector because during the build there was a system property -Dmp.messaging.incoming.channel-stock-price.connector=smallrye-kafka.

I could read the situation wrong. I think this is not a bug, but I'd like to check with you that I didn't miss something. Do you read the change of behavior same as I do and is it expected?

[radcortez]

Env variable matching is a bit smarter. Before it was blindly trying to match the expected dotted name. Now, it uses the actual container types to generate the expected name.

[michalvavrik]

Alright, thanks.