Modified workflow to create new direct-query directory for async-query-core and add last commit_id to the metadata file

[ahkcs]

Overview

This workflow automates the process of building and publishing OpenSearch artifacts to a Maven repository. Specifically, it builds the async-query-core shadow JAR and publishes it to the Sonatype snapshots repository under the org.opensearch.direct-query group ID.

Key Operations

Shadow JAR Building and Processing

Maven Repository Publishing

Employs the standard OpenSearch publish-snapshot.sh script to:
Upload the shadow JAR and associated files
Generate and upload Maven metadata files

Downloads the generated Maven metadata XML file
Adds the commit ID to the metadata for improved traceability
Re-uploads the enhanced metadata file

Also publishes the standard plugin ZIP package using the Gradle task: publishPluginZipPublicationToSnapshotsRepository
Output

The workflow publishes the shadow JAR and all required metadata to: https://aws.oss.sonatype.org/content/repositories/snapshots/org/opensearch/direct-query/{VERSION}-SNAPSHOT/

The published JAR can then be used as a dependency in other projects that need the functionality provided by the async-query-core module.

[noCharger]

I dont think the current shadow jar building depends on opensearch version, can you double check?

[ahkcs]

I think the version is used to name the directory that we upload our jar to

[noCharger]

let's decouple it from opensearch release