Support numpy 2.x.x

• added support for numpy 2.x.x
• fixed depreciation warnings for np.bool
• no change in API

Support numpy 2.x.x

• added support for numpy 2.x.x
• no change in API

v1.1.1

bug fixes for __init__ and __del__ functions

v1.1.0

new property 'is_valid' to check if the mmap of the numpy array is valid

v1.0.1

• added only documentation

v1.0.0

First stable release

• changes API including the class name from ndsharray to NdShArray
• added unit tests
• added documentation on readthedocs
• it is now possible to change shape/dimension/dtype of a numpy array after creation of the ndsharray instance: this will lead to a silently re-creation of the shared memory using mmap. a re-creation can be checked with the property ndarray_mmap_name

v0.2.0

Features:

• now supporting Linux/MacOS
• no restrictions on machines (ARM and AMD 32bit/64bit support)
• small change in API: for Linux/MacOS operation systems information about read ('r') or write ('w') is needed before sharing the array
• more precise timer used to detect write/read time

v0.1.0

Initial Release:

• supporting only Windows OS and amd64 machines
• python >= 3.6
• tested on python 3.6, 3.8 Windows amd64

To Dos:

• add Linux support
• add shared_memory backend for python >= 3.8
• add unit tests
• add documentation