msl.examples.loadlib.cpp64 module

Communicates with cpp_lib32 via the Cpp32 class.

Example of a module that can be executed within a 64-bit Python interpreter which can communicate with a 32-bit library, cpp_lib32, that is hosted by a 32-bit Python server, cpp32. A 64-bit process cannot load a 32-bit library and therefore inter-process communication is used to interact with a 32-bit library from a 64-bit process.

Cpp64 is the 64-bit client and Cpp32 is the 32-bit server for inter-process communication.

class msl.examples.loadlib.cpp64.Cpp64[source]

Bases: Client64

Communicates with a 32-bit C++ library, cpp_lib32.

This class demonstrates how to communicate with a 32-bit C++ library if an instance of this class is created within a 64-bit Python interpreter.

Base class for communicating with a 32-bit library from 64-bit Python.

Starts a 32-bit server, Server32, to host a Python class that is a wrapper around a 32-bit library. Client64 runs within a 64-bit Python interpreter, and it sends a request to the server which calls the 32-bit library to execute the request. The server then provides a response back to the client.

Changed in version 0.6: Added the rpc_timeout argument.

Changed in version 0.8: Added the protocol argument and the default quiet value became None.

Changed in version 0.10: Added the server32_dir argument.

Parameters:

  • module32 (str) – The name of the Python module that is to be imported by the 32-bit server.

  • host (str, optional) – The address of the 32-bit server. Default is '127.0.0.1'.

  • port (int, optional) – The port to open on the 32-bit server. Default is None, which means to automatically find a port that is available.

  • timeout (float, optional) – The maximum number of seconds to wait to establish a connection to the 32-bit server. Default is 10 seconds.

  • quiet (bool, optional) – This keyword argument is no longer used and will be removed in a future release.

  • append_sys_path (str or list of str, optional) – Append path(s) to the 32-bit server’s sys.path variable. The value of sys.path from the 64-bit process is automatically included, i.e., sys.path(32bit) = sys.path(64bit) + append_sys_path.

  • append_environ_path (str or list of str, optional) – Append path(s) to the 32-bit server’s os.environ['PATH'] variable. This can be useful if the library that is being loaded requires additional libraries that must be available on PATH.

  • rpc_timeout (float, optional) – The maximum number of seconds to wait for a response from the 32-bit server. The RPC timeout value is used for all requests from the server. If you want different requests to have different timeout values then you will need to implement custom timeout handling for each method on the server. Default is None, which means to use the default timeout value used by the socket module (which is to wait forever).

  • protocol (int, optional) – The pickle protocol to use. If not specified then determines the value to use based on the version of Python that the Client64 is running in.

  • server32_dir (str, optional) – The directory where the frozen 32-bit server is located.

  • **kwargs – All additional keyword arguments are passed to the Server32 subclass. The data type of each value is not preserved. It will be a string at the constructor of the Server32 subclass.

Note

If module32 is not located in the current working directory then you must either specify the full path to module32 or you can specify the folder where module32 is located by passing a value to the append_sys_path parameter. Using the append_sys_path option also allows for any other modules that module32 may depend on to also be included in sys.path so that those modules can be imported when module32 is imported.

Raises:

  • ConnectionTimeoutError – If the connection to the 32-bit server cannot be established.

  • OSError – If the frozen executable cannot be found.

  • TypeError – If the data type of append_sys_path or append_environ_path is invalid.

add(a, b)[source]

Add two integers.

See the corresponding 32-bit add() method.

Parameters:

  • a (int) – The first integer.

  • b (int) – The second integer.

Returns:

int – The sum of a and b.

subtract(a, b)[source]

Subtract two floating-point numbers (‘float’ refers to the C++ data type).

See the corresponding 32-bit subtract() method.

Parameters:

  • a (float) – The first floating-point number.

  • b (float) – The second floating-point number.

Returns:

float – The difference between a and b.

add_or_subtract(a, b, do_addition)[source]

Add or subtract two floating-point numbers (‘double’ refers to the C++ data type).

See the corresponding 32-bit add_or_subtract() method.

Parameters:

  • a (float) – The first floating-point number.

  • b (float) – The second floating-point number.

  • do_addition (bool) – Whether to add the numbers.

Returns:

float – Either a + b if do_addition is True else a - b.

scalar_multiply(a, xin)[source]

Multiply each element in an array by a number.

See the corresponding 32-bit scalar_multiply() method.

Parameters:

  • a (float) – The scalar value.

  • xin (list of float) – The array to modify.

Returns:

list of float – A new array with each element in xin multiplied by a.

reverse_string_v1(original)[source]

Reverse a string (version 1).

In this method Python allocates the memory for the reversed string and passes the string to C++.

See the corresponding 32-bit reverse_string_v1() method.

Parameters:

original (str) – The original string.

Returns:

str – The string reversed.

reverse_string_v2(original)[source]

Reverse a string (version 2).

In this method C++ allocates the memory for the reversed string and passes the string to Python.

See the corresponding 32-bit reverse_string_v2() method.

Parameters:

original (str) – The original string.

Returns:

str – The string reversed.

distance_4_points(points)[source]

Calculates the total distance connecting 4 Point’s.

See the corresponding 32-bit distance_4_points() method.

Attention

This method does not work with if Cpp64 is running in Python 2. You would have to create the FourPoints object in the 32-bit version of distance_4_points() because there are issues using the pickle module between different major version numbers of Python for ctypes objects.

Parameters:

points (FourPoints) – Since points is a struct that is a fixed size we can pass the ctypes.Structure object directly from 64-bit Python to the 32-bit Python. The ctypes module on the 32-bit server can load the pickle’d ctypes.Structure.

Returns:

float – The total distance connecting the 4 Point’s.

circumference(radius, n)[source]

Estimates the circumference of a circle.

This method calls the distance_n_points function in cpp_lib32.

See the corresponding 32-bit circumference() method.

Parameters:

  • radius (float) – The radius of the circle.

  • n (int) – The number of points to use to estimate the circumference.

Returns:

float – The estimated circumference of the circle.