feat(conversations): add conversation and model management APIs

[tharropoulos]

Rationale

This PR introduces conversation management capabilities to the Typesense Java client, allowing developers to integrate LLM-powered conversations into their applications.

This addition enables Java developers to:

• Configure and manage LLM models through Typesense
• Store and retrieve conversation histories
• Set up model-specific parameters like context windows and system prompts
• Integrate with OpenAI and other LLM providers

Changes

Added Features

1. New Classes for Conversation Management:

   • ConversationModel: Handles individual model operations (CRUD)
   • ConversationModels: Manages model collections
   • Conversation: Manages individual conversations
   • Conversations: Handles conversation collections

2. New Schema Classes:

   • ConversationSchema: Defines conversation structure
   • ConversationModelCreateSchema: Model creation parameters
   • ConversationModelUpdateSchema: Model update parameters
   • ConversationDeleteSchema: Deletion response structure

Code Changes

1. In Client.java:

   • Added conversation management support
   • Implemented new methods: conversations() and conversations(String id)
   • Added conversation-related fields and initialization

2. In Helper.java:

   • Added createConversationCollection() for test setup
   • Added createTestConversationModel() for testing
   • Updated teardown() to clean up conversation resources

3. In build.gradle:

   • Fixed JUnit platform configuration for CI environment

Test Updates

1. Added ConversationModelsTest.java:
   • Test coverage for model CRUD operations
   • Integration tests with OpenAI GPT-3.5
   • Environment variable handling for API keys

PR Checklist

• I have read and signed the Contributor License Agreement.