[Test][E2E] Bound Pulsar sink IT receive wait

[DanielLeens]

What does this PR do?

Bound PulsarSinkIT consumer reads by a total wait window and poll timeout so the test fails fast instead of hanging the whole matrix job.

Why is the change needed?

PR #11136 surfaced a cancelled updated-modules-integration-test-part-1 job that is unrelated to its functional change. The actionable failure comes from PulsarSinkIT, which currently blocks forever on consumer.receive() when the expected 10 sink records do not arrive in time.

How does it work?

• use a bounded receive loop driven by the expected record count instead of a fixed number of blocking calls
• poll Pulsar with a timeout so the test never leaves CI stuck in an infinite wait
• keep the assertion on 10 output records and log a clear warning when the timeout is exhausted

How was this PR tested?

• ./mvnw spotless:apply
• ./mvnw -pl seatunnel-e2e/seatunnel-connector-v2-e2e/connector-pulsar-e2e -DskipTests test-compile
• attempted real runtime validation: ./mvnw -pl seatunnel-e2e/seatunnel-connector-v2-e2e/connector-pulsar-e2e -Dtest=PulsarSinkIT -Dsurefire.failIfNoSpecifiedTests=false test
  • the Testcontainers run started, but local Docker Desktop crashed apachepulsar/pulsar:2.3.1 with a container-side JDK 8 SIGSEGV before the SeaTunnel assertions completed
• not revalidated in this run: seatunnel-engine-ui (unchanged and outside the affected module path)