better / simpler install defaults

also enhance external project not rebuilding each time

cmake bugfix: build if template change

• Cmake build bugfix: detect template change and regenerate
• update cmake presets and use them in CI

auto-find HDF5 from installed package

install: find HDF5 automatically

packages using h5fortran install no longer need to find HDF5 themselves

install Find*.cmake

This facilitates finding HDF5 robustly, as we believe our FindHDF5 is better suited to h5fortran use cases than the CMake factory FindHDF5.cmake

CI enhancements

Use CDash as a first-class CI use case.
Eliminated intrinsic-shadow warning.

script fixes

• FindHDF5.cmake: fix detection of self-built HDF5 on Unix-like OS
• build_hdf5.py: bugfix and allow building git revision or download zip

allow HDF5 install to be outside build dir

This allows user to specify HDF5_ROOT and ZLIB_ROOT to say ~/lib_gcc so that the HDF5 library can be built and installed just once.

int64 read 0d..3d write 0d..7d

add ability to read/write int64 data. The HDF5 Fortran interface doesn't currently have an H5T_NATIVE_INT64 so we assume the int64 data on disk is H5T_STD_LE64.

If needed, we can add read int64 for 4d..7d in the future.

improve packaging, generate h5fortran.pc

• generate h5fortran.pc for pkg-config
• improve packaging
• export targets to build dir

read: add input dataset validation and buffering

with GCC-10 and -O3 on Windows, under certain conditions it was observed that reading datasets with implied type coercion would fail by giving random data, changing run to run of the same program/data. This was observed on scalars only.
To solve this issue and add general robustness to reading, we now get the dataset type from disk and compare it with the variable datatype, raising an error if the variable type and size do not match the dataset type and size.