MATLAB: Are C++ shared library targets supported with MATLAB Compiler 3.0 (R13)

ccompilerdocumentationerrorlibraryMATLAB Compilersharedstatic

The MATLAB Compiler 3.0 (R13) documentation for MCC states:

 -B cpplib (C++ Library).   Produce a C++ library. 
 The -B cpplib option is equivalent to the series of 
 options -B csharedlib:
 <shared_library_name> 
 -L cpp -T compile:lib

Does this indicate that C++ shared library targets are supported?

Best Answer

This has been verified as an error within the documentation for the MATLAB Compiler within the section for MCC. C++ shared library targets are not supported with MATLAB Compiler 3.0 (R13).

Related Solutions

MATLAB: Are graphics supported in MATLAB Compiler generated shared libraries

As of MATLAB Compiler 4.0 (R14), graphics are officially supported in Compiler generated shared libraries. M-functions that use graphics can be compiled into a C shared library using the "-B csharedlib" option, just like compiling a non-graphical M-function into a C shared library. If you are using a previous version of MATLAB Compiler, see the following:

For MATLAB 6.5 (R13), The MATLAB C/C++ Graphics Library is only officially supported for use in MATLAB Compiler generated executables.

Although generated shared libraries that use graphics may in some cases function without error, the MATLAB C/C++ Graphics Library has not been fully tested for use in a MATLAB Compiler generated shared library and is not guaranteed to work. We are unable to provide assistance for cases where such a library does not work.

Note that MATLAB COM Builder allows you to generate shared libraries which include MATLAB graphics routines in the form of a COM (Common Object Model) object.

MATLAB: How to pass strings into a C++ shared library built using MATLAB Compiler 4.5 (R2006b)

This example passes string variables into and out of a C++ shared library built with MATLAB Compiler 4.5 (R2006b). It is based on the C++ shared library example in the documentation

 web([docroot,'/toolbox/compiler/f2-995712.html'])

1. Building the shared library

Download the following attached files your work directory, using the links at the bottom of this page:

 revstr.m
 revdriver.cpp

Note: revdriver.cpp contains the driver application's main function.

2. Compiling the library

First, choose your C++ compiler:

mbuild -setup

To create the shared library, execute the following command:

mcc -B cpplib:librevp revstr.m -v

The arguments to the MCC command are explained in the C++ shared library example.

3. Compiling the driver application

To compile the driver code (revdriver.c) using the selected C/C++ compiler, execute the following platform-specific MBUILD command

mbuild revdriver.cpp librevp.lib      (Windows)
mbuild revdriver.cpp -L. -lrevp -I.  (UNIX)

This generates a driver application, revdriver.exe, on Windows, and revdriver, on UNIX.

4. Running the driver application

To execute the standalone application, add the directory containing the shared library to your dynamic library path. Also update the dynamic library path for your platform by following the instructions in the section 'Developing and Testing Components on a Development Machine' of the 'Getting Started' chapter.

Execute the driver application from the system command prompt by typing the application name:

 revdriver

The results are displayed as:

 The value of the original string is:
 This is a string...
 The value of the modified string is:
 ...gnirts a si sihT
 Viewed as a C++ string, the value is:
 ...gnirts a si sihT

Best Answer

Related Solutions

MATLAB: Are graphics supported in MATLAB Compiler generated shared libraries

MATLAB: How to pass strings into a C++ shared library built using MATLAB Compiler 4.5 (R2006b)

Related Question