Release 14.1#54
Merged
Merged
Conversation
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The generator-interfaces build downloaded XMLSchema.xsd, XMLSchema.dtd, and datatypes.dtd live from www.w3.org on every build to generate the w3c.xsd JAXB classes. These are frozen 2009 standards, and the live fetch made the build non-hermetic and prone to HTTP 429 rate-limiting from W3C (which broke CI). Commit the three files under src/main/xsd-ref/, point the process-w3c-xsd xjc execution at the local copy, and remove the wagon-maven-plugin download executions. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
For a synchronous INVOKE with a registered listener, the same lazily decoded message body is read by two threads: the consumer thread woken by the interaction handler and the dispatcher thread delivering the same message to the listener. decodeMessageBody() was not synchronized and set decodedBody=true before allocating messageParts, so the losing thread saw "already decoded", read the still-null messageParts, and failed with an NPE far from the cause. The window only opens for a non-empty ACK body, which is why it went unnoticed. Make the method synchronized. Also stop swallowing decode failures: a NotFoundException left messageParts null and readers would NPE later with no hint of the cause. Rethrow it as MALException like the sibling branch, and make repeated reads after a failed decode fail with a MALException instead of an NPE. Verified against the NMF e2e testbed (a sync install invoke with a populated ACK reproduced the NPE deterministically on the unfixed code and passes with this fix) and the full MAL testbed suite. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Rethrowing the NotFoundException as MALException changed the behavior of a long-standing path: callers above may push the error back to consumers or brokers that previously carried on after the logged warning. Keep the original swallowing behavior; the repeated-read guard still reports a descriptive MALException instead of an NPE when a previous decode attempt has failed. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
loadConsumerProperties() re-read and re-applied consumer.properties on every call. Consumers such as the CTT invoke it once per connection (via NMFConsumer.init() and retrieveProvidersFromDirectory()), so the file was loaded repeatedly during startup, producing a stream of "Reloading properties: consumer.properties" log lines. Guard it with a dedicated "ConsumerPropertiesLoadedFlag" so the file is loaded at most once. The flag is deliberately independent from the provider's "PropertiesLoadedFlag": an application acting as both a consumer and a provider must still load both property files. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The repeated-read guard added in 277f7c8 throws a MALException when a body is read with decodedBody=true but messageParts=null, on the assumption that this only happens after a decode failure. That assumption was wrong: an empty body is passed to the constructor as a null messageParts array, so every argument-less request and every empty acknowledgement hit the guard, and getElementCount() logged it at SEVERE during encoding on both the provider and consumer sides. Guarantee the invariant "decodedBody == true implies messageParts != null" at both assignment sites: - the constructor coalesces a null messageParts to an empty array; - the swallowed-NotFoundException branch (the tolerated operation-not- found path restored in 07d8f6e) degrades to an empty body instead of leaving messageParts null. Empty bodies now encode as zero-part bodies as they did before 277f7c8, the SEVERE spam is gone, and the guard remains as defensive insurance for a genuinely inconsistent state. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
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
1 participant
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.
Release of version 14.1.
Changes since 14.0
To be Squash and Merged, per RELEASING.md.
🤖 Generated with Claude Code