I wanted to cross compile Leptonica on Linux, Fedora 30 in my case, using the MXE cross compile libraries. To do this I modified the makefile.static file in the leptonica/src directory as suggested on the Leptonica website.
My MXE installation is in '/opt/mxe'. If you have put it elsewhere then you will need to modify the paths appropriately.
I first ran ./make-for-local which copies src/makefile.static to src/makefile. Then I modified 'makefile' to allow the use of a COMPILER flag.
So if you want to use these files you should copy makefile into the leptonica/src directory 'cd src' to move to that directory and run make with one of the following commands.
All versions create a library called 'liblept.a'.
- 'make' This will create a no debug static library using your local gcc in leptonica/nodebug.
- 'make COMPILER=GCC' This is exactly the same as above as GCC is the default.
- 'make DEBUG=yes debug' This will create a debug static library sing your local gcc in leptonica/debug.
- 'make COMPILER=GCC DEBUG=yes debug' This is exactly the same as above as GCC is the default.
- 'make COMPILER=MXE_GCC_STATIC' This will create a no debug static library using the MXE static gcc in leptonica/nodebug.
- 'make COMPILER=MXE_GCC_STATIC DEBUG=yes debug' This will create a debug static library using the MXE static gcc in leptonica/nodebug.
There is a COMPILER=MXE_GCC_SHARED flag but that doesn't work at the moment and as I don't want the shared library at the moment I haven't tried to fault find it.
You can chain the methods together and so long as you don't 'make clean, all the files will be created in separate directories.
make; make DEBUG=yes debug; make COMPILER=MXE_GCC_STATIC; make COMPILER=MXE_GCC_STATIC DEBUG=yes debug
will create both unix and mingw debug and nodebug static libraries.
I am placing all my created library files in a central location under ~/workspace/lib.
lib lib/unix lib/unix/lib # 'liblept.a' and 'libleptd.a' (the debug version was renamed to unix style) lib/unix/include lib/mingw64 lib/mingw64/release lib/mingw64/release/lib lib/mingw64/release/include lib/mingw64/debug lib/mingw64/debug/lib lib/mingw64/debug/include
the shell script install.sh makes this directory structure and copies the various '.h' files and the ".a" library files into this structure.
I am certain that there is a more elegant way to do this but it works and I don't have the time to chase down faults and/or a better method. Maybe later if I get the time.
All this is placed under the MIT license so if it is of any use to you, then use it.