This pull request introduces several significant improvements and updates across the build system, CI workflows, and the core implementation of quad-precision constants for the quaddtype package. The main highlights are the addition of a new quad-precision constants header, improved handling of build dependencies (especially for NumPy), and updates to CI to support newer Python and OS versions.

Summary of the most important changes:

Quad-Precision Constants Implementation

• Added a new header constants.hpp defining quad-precision floating-point and integer constants, as well as a function for querying constants by name, leveraging SLEEF library macros and types.
• Integrated the new constants into the quad-precision dtype implementation, including a new quadprec_get_constant function and registration with NumPy's DType API. [1] [2] [3] [4]
• Included the new header in the build system and ensured the relevant source file is compiled.

Build and Dependency Management

• Updated all CI and build workflows to install NumPy from nightly wheels or source (main branch), ensuring compatibility with the latest Python versions and features. This includes using --no-build-isolation and --break-system-packages where appropriate. [1] [2] [3] [4] [5] [6] [7] [8]
• Added version checks and informative output for the NumPy version used during the build process.

Continuous Integration and Platform Support

• Updated base images and Python versions in CI workflows to Ubuntu 24.04 and Python 3.11, and dropped support for Python 3.10 in wheel builds and CI. [1] [2] [3] [4] [5] [6]
• Improved separation of build and test dependencies in the big endian workflow, installing test requirements only after the main package is built.

Type Stubs

• Cleaned up type ignore comments for reflected operators in the QuadPrecision type stub.

closes #164