Introduce retry with backoff

[jmuziki]

This pull request includes several changes to the invoke method in the src/fabric_cicd/_common/_fabric_endpoint.py file, focusing on improving error handling and retry logic for API requests. The most important changes include the addition of a new _handle_retry method, enhanced error messaging, and improved handling of API throttling and long-running operations.

Improvements to error handling and retry logic:

• Added a new _handle_retry method to handle retry logic with exponential backoff based on the response headers. This method is called when the Retry-After header is present in the response.
• Enhanced error messaging for failed operations and deployment attempts, improving readability and clarity. [1] [2]

Handling of API throttling and long-running operations:

• Improved handling of API throttling by using the _handle_retry method when the Retry-After header is present in the response. This change ensures that the retry logic is consistent and robust.
• Added logic to handle retries for long-running operations when the Retry-After header is present in the response. This change ensures that the system waits appropriately before retrying the operation.

Minor improvements:

• Fixed a typo in a log message to improve clarity.