Welcome to PyCLibrary’s documentation!¶
PyCLibrary tries to make wrapping dynamic librariries in python less cumbersome and more user friendly.
The idea is that most of the things needed such as constant values and function signatures are already presents in the headers files of the library (which are usually accessible as they are needed for using the library in C). So better to use them than copy everything by hand.
PyCLibrary includes 1) a pure-python C parser and 2) an automation library that uses C header file definitions to simplify the use of c bindings. The C parser currently processes all macros, typedefs, structs, unions, enums, function prototypes, and global variable declarations, and can evaluate typedefs down to their fundamental C types + pointers/arrays/function signatures. Pyclibrary can automatically build c structs/unions and perform type conversions when calling functions via cdll/windll.
PyCLibrary tries to present a ffi agnostic API to allow using different bindings. For the time being only the ctypes based backend is implemented but a cffi backend should be possible to implement (the rational for it would be that the CParser can be used on raw header files which are not always well supported by the cffi parser).
However if you need to manipulate the C object coming back from the library which cannot simply be mapped to Python object your code will most likely not be backend independent so it is discouraged to try to switch between backends.
PyCLibrary supports Python 2.7 and 3.3+
How to set up PyCLibrary and make your first step with it.
More references on PyCLibrary internals.
Some questions that might have occurred to others too.
When all else fails, consult the API docs to find the answer you need. The API docs also include convenient links to the most definitive PyCLibrary documentation: the source.