Add e2e test for S3 destination processing pipeline

[PuneetPunamiya]

• Tests the complete flow from source S3 bucket through document processing, chunking, and vector embeddings generation to destination S3 bucket

[gemini-code-assist]

Code Review

This pull request introduces a new end-to-end test, TestS3Destination, which validates the unstructured data processing pipeline using S3 for both source and destination. The feedback focuses on improving the robustness and cleanliness of the test code. Key suggestions include avoiding hardcoded SQS URLs by capturing them from the creation response, using t.Fatal for setup failures to prevent invalid test states, removing redundant client initializations, and ensuring the teardown logic comprehensively cleans up all created resources, including the data storage bucket and SQS queue.

[gshikhar2021]

@PuneetPunamiya , After rebasing you haven't rename the unstructureddataproduct to unstructureddatapipeline at line 942 and 1012 as function name is updated , please take a look of that

• also the above comment needs to be addressed imo, by directly deleting the bucket instead of cleaning and then deleting
• just a small improvement as per my suggestion ,can we use the same buckets instead of creating new one , so basically you can move the bucket objects deletion code into the cleanup logic of Test1 and when the Test2 will start it will get fresh buckets and eventually in the cleanup of Test2 you can drop the buckets, wdyt ??