Upgrade to use C++20.

Closes: USTC-KnowledgeComputingLab/qmb#40
Closes: USTC-KnowledgeComputingLab/qmb#36

Author: hzhangxyz

qmb/_hamiltonian.cpp

@@ -37,9 +37,9 @@ auto prepare(py::dict hamiltonian) {
     auto coef = torch::empty({term_number, 2}, torch::TensorOptions().dtype(torch::kFloat64).device(torch::kCPU));
     // No need to initialize
 
-    auto site_accessor = site.template accessor<std::int16_t, 2>();
-    auto kind_accessor = kind.template accessor<std::uint8_t, 2>();
-    auto coef_accessor = coef.template accessor<double, 2>();
+    auto site_accessor = site.accessor<std::int16_t, 2>();
+    auto kind_accessor = kind.accessor<std::uint8_t, 2>();
+    auto coef_accessor = coef.accessor<double, 2>();
 
     std::int64_t index = 0;
     for (auto& item : hamiltonian) {

qmb/hamiltonian.py

@@ -31,8 +31,8 @@ def _load_module(cls, n_qubytes: int = 0, particle_cut: int = 0) -> object:
                     f"{folder}/_hamiltonian_cuda.cu",
                 ],
                 is_python_module=n_qubytes == 0,
-                extra_cflags=["-O3", "-ffast-math", "-march=native", f"-DN_QUBYTES={n_qubytes}", f"-DPARTICLE_CUT={particle_cut}"],
-                extra_cuda_cflags=["-O3", "--use_fast_math", f"-DN_QUBYTES={n_qubytes}", f"-DPARTICLE_CUT={particle_cut}"],
+                extra_cflags=["-O3", "-ffast-math", "-march=native", f"-DN_QUBYTES={n_qubytes}", f"-DPARTICLE_CUT={particle_cut}", "-std=c++20"],
+                extra_cuda_cflags=["-O3", "--use_fast_math", f"-DN_QUBYTES={n_qubytes}", f"-DPARTICLE_CUT={particle_cut}", "-std=c++20"],
                 build_directory=build_directory,
             )
         if n_qubytes == 0:  # pylint: disable=no-else-return