Implement rate-limiting for OpenAI API calls

[penguineer]

This pull request introduces a rate-limiting mechanism to handle API rate limits more gracefully. The most important changes include adding a RateLimitGate component, modifying the ChatRequestHandler to use this new component, and updating the RateLimitException handling to consider an additional header.

This works towards #8

Rate Limiting Mechanism:

• src/main/java/com/penguineering/hareairis/rmq/RateLimitGate.java: Added the RateLimitGate component to manage rate limits by waiting for the next available time before executing a protected call.

Integration with Chat Request Handler:

• src/main/java/com/penguineering/hareairis/rmq/ChatRequestHandler.java: Updated the ChatRequestHandler to include the RateLimitGate and use it in the onMessage method to handle chat requests within rate limits. [1] [2] [3]

Exception Handling:

• src/main/java/com/penguineering/hareairis/ai/RateLimitException.java: Enhanced the RateLimitException handling to check for the x-ratelimit-timeremaining header if the Retry-After header is not present.

Configuration Changes:

• src/main/java/com/penguineering/hareairis/rmq/RabbitMQConfig.java: Updated the RabbitMQ configuration to ensure the RateLimitGate bean is initialized before the chatRequestsContainer bean. [1] [2]

Copilot reviewed 4 out of 4 changed files in this pull request and generated no suggestions.

Comments skipped due to low confidence (1)

src/main/java/com/penguineering/hareairis/rmq/ChatRequestHandler.java:35

• [nitpick] Catching a generic Exception and throwing a RuntimeException in 'serializeChatError' is not ideal. Consider handling specific exceptions.

return chatEx.map(ChatError::new).map(err -> { try { return objectMapper.writeValueAsString(err); } catch (JsonProcessingException ex) { throw new RuntimeException(ex); } });