Improve deserialisation and payment contexts updates

[armando-rodriguez-cko]

Description

This PR improves deserialisation processes and updates key features in the payment contexts and test implementations, ensuring enhanced flexibility and robustness when handling diverse input formats and data structures.

Changes Made

Improved Deserialisation:

• Enhanced the GsonSerializer to handle various date formats:
  • ISO-8601 Instants (yyyy-MM-ddTHH:mm:ssZ).
  • Numeric date formats like yyyyMMdd.
  • Standard date formats like yyyy-MM-dd.
• Incorporated fallback mechanisms to gracefully parse dates using multiple predefined formats, ensuring no crashes occur when invalid or unexpected date formats are received.

Updates to Payment Contexts:

• Adjusted PaymentContextsAirlineData to handle tickets as a single entity rather than a list, improving clarity and alignment with processing expectations.

Test Enhancements:

• Added unit tests to validate deserialisation for all supported date formats:
  • Covered cases with numeric formats (yyyyMMdd).
  • Handled edge cases like missing or invalid date values.
• Extended tests for payment contexts:
  • Included scenarios for Klarna and PayPal responses in mock data.
  • Validated serialisation and deserialisation for payment context objects.

Mock Data Updates:

• Updated JSON mocks to align with the latest payment context changes, ensuring consistency across tests.
• Added more realistic test cases to simulate production-like scenarios, including complex payment processing details.

Impact

• Improves the SDK's robustness and flexibility by handling diverse and non-standard input formats.
• Enhances reliability in deserialisation, reducing the likelihood of errors during runtime.
• Strengthens test coverage to ensure the system remains stable under edge cases.

Additional Notes

• No breaking changes were introduced, maintaining full backward compatibility with existing implementations.
• These updates ensure a smoother developer experience by explicitly handling edge cases and improving diagnostics for deserialisation errors.