Add a Google Cloud Storage file source
Implemented SeaTunnel's Google Cloud Storage source by integrating the `gs` filesystem with the existing multiple-table file ingestion framework.
apache/seatunnel · #11967
New cloud storage connector
SeaTunnel can now ingest files from Google Cloud Storage with schema discovery, splitting, continuous discovery, and multiple authentication paths.
Problem
SeaTunnel's file source framework supported formats, schema discovery, splitting, and continuous discovery, but lacked a packaged GCS filesystem source with validated bucket configuration, credential handling, and distribution registration.
Approach
Added `gs://` bucket support with separately configured object paths, Application Default Credentials, optional service-account JSON keys, additional `fs.gs.*` properties, filesystem service registration, plugin packaging, license metadata, E2E coverage, and English and Chinese documentation.
Impact and scope
- Adds native Google Cloud Storage ingestion while reusing SeaTunnel's mature file-source behavior.
- Supports both default cloud credentials and explicit service accounts with deterministic configuration precedence.
- Enables bounded and continuously discovered file pipelines without changing existing file connectors.
Validation
- Added tests for bucket validation, credential modes, property precedence, factory options, and Hadoop filesystem resolution.
- Eight focused tests and repository verification passed on Java 8 and Java 11; the final Apache Build also passed without claiming a credentialed live-cloud read.
- Four authored commits and the GitHub-verified merge credit Goutam Adwant; upstream approval landed the feature.