RFQ: Add maxRetrySeconds configuration and RetryTooLong exception

Summary

This PR implements the maxRetrySeconds configuration parameter and RetryTooLong exception to limit retry wait times for the Zscaler SDK, addressing ticket SZ-1.

Changes Made

1. Configuration (zscaler/config/config_setter.py)

• Added maxRetrySeconds parameter to the rateLimit configuration section
• Default value is 0 (meaning unlimited retry time)
• Added to both _DEFAULT_CONFIG and _apply_default_values method

2. Exception Class (zscaler/exceptions/exceptions.py)

• Created new RetryTooLong exception class
• Includes retry_seconds and max_seconds properties
• Descriptive error message showing both actual and maximum retry times

3. Request Executor (zscaler/request_executor.py)

• Added validation in constructor to ensure maxRetrySeconds >= 0
• Modified fire_request_helper method to check backoff time against limit
• Logs warning before throwing RetryTooLong exception
• Only applies limit when maxRetrySeconds > 0 (0 means unlimited)
• Check applies to all retryable status codes (408, 409, 412, 429, 500, 502, 503, 504)

4. Module Exports (zscaler/__init__.py)

• Exported RetryTooLong exception for easy import

Usage Example

from zscaler import ZscalerClient, RetryTooLong

with ZscalerClient(
    config={
        "client": {
            "rateLimit": {
                "maxRetrySeconds": 50
            }
        }
    }
) as client:
    try:
        results = client.zcc.devices.list_device()
    except RetryTooLong:
        print("Retry time is too long.")

Testing

• ✅ All configuration changes verified
• ✅ Exception class properties and message tested
• ✅ RequestExecutor validation tested
• ✅ Import functionality verified
• ✅ Build verification passed with poetry build

Implementation Details

• The feature only activates when maxRetrySeconds > 0
• Warning is logged before throwing the exception
• Applies to all retryable HTTP status codes, not just 429
• Maintains backward compatibility (default value 0 = unlimited)

Jira Ticket: SZ-1
Link to Devin run: https://app.devin.ai/sessions/d0e3dce18ac34f3badb1e989d3d98256
Requested by: Samir Chaudhry (samir@cognition.ai)