[FLINK-39588][flink-table][connector/datagen] Migrate the DataGen table connector off deprecated source APIs

[Dennis-Mircea]

What is the purpose of the change

This pull request migrates the built-in 'connector' = 'datagen' table source (DataGenTableSource in flink-table-api-java-bridge) off the long-deprecated SourceFunctionProvider and the legacy SourceFunction-based DataGeneratorSource / RandomGenerator / SequenceGenerator hierarchy, onto the FLIP-27 source stack (Source API + SourceProvider).

These were the last remaining usages of those deprecated APIs from a built-in table connector and contributed a sizeable chunk of the frozen entries in flink-architecture-tests-production. User-visible behavior of the connector is preserved: same DDL options, same value distributions, same parallelism, same rows-per-second rate limiting, and same termination semantics for bounded sequence fields.

Brief change log

• Replaced SourceFunctionProvider with SourceProvider.of(Source, parallelism) in DataGenTableSource.
• Switched the underlying source to FLIP-27 org.apache.flink.connector.datagen.source.DataGeneratorSource, driven by a composite GeneratorFunction<Long, RowData> and RateLimiterStrategy.perSecond(rowsPerSecond).
• Introduced new RandomGeneratorFunction<T> / SequenceGeneratorFunction<T> types implementing GeneratorFunction<Long, T>, replacing the legacy RandomGenerator / SequenceGenerator hierarchies. Migrated RowDataGenerator, DataGeneratorMapper, DecimalDataRandomGenerator, DataGeneratorContainer, and the random/sequence visitors accordingly.
• Sequence checkpoint state ownership moved from per-generator ListState<Long> deques to the NumberSequenceSource driving the FLIP-27 source. The legacy "halt when any sequence range is exhausted" semantic is preserved via a source-level effective count (min(sequenceFieldsTotalCount) when number-of-rows is unset).
• Added flink-connector-datagen as a dependency of flink-table-api-java-bridge.
• Removed the now-stale entries from flink-architecture-tests-production/archunit-violations/.

Verifying this change

This change added tests and can be verified as follows:

• DataGenTableSourceFactoryTest rewritten to drive the generator function directly via DataGenTableSource#buildRowGenerator() + computeEffectiveCount() (no MiniCluster needed). All 14 existing factory + per-field behavior assertions are retained.
• New test testSequenceProducesEachValueExactlyOnce asserts every value in the requested [start, end] range is produced exactly once.
• DecimalDataRandomGeneratorTest updated for the new map(Long) entry point.
• Architecture tests in flink-architecture-tests-production pass with the legacy-API violation entries removed from the frozen list.
• Existing ITCases that exercise 'connector' = 'datagen' (e.g. DataGeneratorConnectorITCase, SqlGatewayServiceITCase, AbstractMaterializedTableStatementITCase, OpenAIChatModelTest, ProcessTableFunctionTest) continue to pass unchanged.

Does this pull request potentially affect one of the following parts:

• Dependencies (does it add or upgrade a dependency): yes (flink-table-api-java-bridge now depends on flink-connector-datagen)
• The public API, i.e., is any changed class annotated with @Public(Evolving): no
• The serializers: no
• The runtime per-record code paths (performance sensitive): no
• Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: no
• The S3 file system connector: no

Documentation

• Does this pull request introduce a new feature? no
• If yes, how is the feature documented? not applicable

---

Was generative AI tooling used to co-author this PR?

• Yes (please specify the tool below)

Generated-by: [GitHub Copilot]

[flinkbot]

CI report:

• dca055e Azure: SUCCESS

Bot commands

The @flinkbot bot supports the following commands:

• @flinkbot run azure re-run the last Azure build