Refactor Java Code Quality Analyzer

[samvaity]

What does this implement/fix? Explain your changes.

This PR includes changes for Azure Toolkit for IntelliJ's Code Quality Analyzer to integrate rule sets for flagging and making better Azure SDK usage.

Integrated Rule Sets:

1. Storage:

   • Ensure Storage Upload APIs use a length parameter to prevent memory issues.

2. Identity & Security:

   • Avoid hardcoded API keys; use DefaultAzureCredential instead.

3. Asynchronous Programming:

   • Use SyncPoller instead of PollerFlux#getSyncPoller() for simplicity.

4. Messaging (AMQP & Service Bus):

   • Prefer ServiceBusProcessorClient over ServiceBusReceiverAsyncClient.
   • Explicitly disable auto-complete in Service Bus clients to avoid message loss.
   • Optimize Receive Mode & Prefetch Value for efficient processing.
   • Use EventProcessorClient for checkpoint management.

5. General Best Practices:

   • Batch operations instead of single operations in loops.
   • Use recommended authentication methods over connection strings.
   • For Azure OpenAI, prefer getChatCompletions over getCompletions.

Does this close any currently open issues?

Closes open Pull Request #8584

Any relevant logs, screenshots, error output, etc.?

1. Storage Uppload API without length check:

2. DisableAutoComplete check for ServiceBus
   

3. Prefer EventProcessorClient instead of EventHubConsumerAsyncClient

4. Block over Async clients check

5. Use of hardcoded keys in key based Authentication

Any other comments?

…

Has this been tested?

• Tested

[wangmingliang-ms]

I'm not quite sure, do you know if this could run prior to ProjectActivity? If that's the case, there might be a NPE, given that INSTANCE might not have initialized yet.

[samvaity]

I did do a runIde for testing this strategy. It does seem to be working as expected.
Also, from the documentation here, <postStartupActivity/> is executed after project initialization but before the first UI interaction. So, it should be safe relying on the project being fully initialized can safely execute logic.

[wangmingliang-ms]

LGTM