I'm a bit stuck trying to get box2D to compile and I think it's because I don't quite understand how Xcode handles its build directories.
Box2D is folder containing a set of header and source files (in various subdirectories, etc). I've added the Box2D folder to a coco touch static library project in Xcode and when I try to compile I get errors about header files not being found such as <Box2D/Common/b2BlockAllocator.h>.
I found that if I simply include the header with #include "b2BlockAllocator.h" it compiles fine and Xcode actually finds the file.
So I'm a bit stuck here, I'm assuming I need to find a way to get all includes to begin searching from the root project directory and not from the source files location but I'm not sure how I can do that in Xcode...
Any ideas?
<Box2D/Common/b2BlockAllocator.h> should be found in a directory called Box2D/Common. You said that Box2D contains a set of header and source files. Are some of these headers in a subdirectory of Box2D called Common, or are they all directly under Box2D? If it's the latter, then that's your problem.
OK I figured it out!
In the targets settings page look for 'header search paths' and add the root build directory of your project to it.
Related
I have recently downloaded SDL2 and am trying to install it. I created a small test program consisting of a .cpp and a .h file.
I have put SDL2.frameworks inside of /Library/Frameworks and then made sure that it was added to General->Frameworks and Libraries and tried to #include <SDL2/SDL.h>. file not found.
I made sure to go to the build settings->Search Paths->Framework Search Paths to add /Library/Frameworks (it was already listed next to the setting but I readded it to the dropdown). I also added /Library/Frameworks/SDL2.framework to the Header Search Paths. file not found.
I tried to include <SDL2/SDL.h>, <SDL/SDL.h>, <SDL.h>, "SDL2/SDL.h", "SDL/SDL.h" and "SDL.h". none of these files were found.
For the header search path, it should be set to FRAMEWORK_DIRCTORY/SDL2.framework/Headers/, or you can use FRAMEWORK_DIRCTORY/SDL2.framework and turn on recursive search.
Then you should be able to use it with just #include <SDL.h>.
Edit:
After some investigation, seems like my original answer was more of a workaround. The proper way of including SDL library in your code should be using just #include <SDL2/SDL.h>, however it wasn't working.
The reason of that is when compiling the code, Xcode would attempt to copy the library to the product folder, so the compiled executable can have easy access to it. However, for some reason, Xcode copied the framework without the header folder (the reason is probably that Xcode is moving towards using Swift only, which doesn't really have a "header" thing).
By default, when you try to run the code, and it sees there is a SDL2.framework folder located in the product folder, it would use that framework, even if it doesn't have the header folder located in it. But since it doesn't actually have the header folder in it, it doesn't actually run.
To solve it, the easiest way is to remove it from the Targets/Build Phases/Embed Frameworks.
By default when you add a framework to your project, it would also be added here. By removing it from the Embed Frameworks, it won't copy the framework to product folder. And it will only attempt to search SDL framework from whatever you have put in the Build Settings/Framework Search Path.
In the same time, you won't need to have anything for Build Settings/Header Search Path, as it will look for Headers folders in your frameworks first.
And with you code, you can just use #include <SDL2/SDL.h>.
Even better, you can add the framework to Build Phases/Copy Files, and set the Destination to Frameworks, empty the Subpath, and potentially uncheck Copy only when installing, for better cross device usabilities.
I'm developing a code which uses Easylogging++ as the underlying logging library. Recently, I wanted to update the library since it has some high visibility / high impact bugs and I found out that the library is divided into two files (.cc and .h). This new structure needs inclusion of the .cc file in the build string alongside the main program code.
I'm using Eclipse to develop the project and generate the make files to build the project. I need to tell Eclipse (Oxygen.1) that it needs to compile the .cc file alongside the main file while building the project, however I was unable to do so. Any help is greatly appreciated.
It's easier than I thought. Eclipse's managed build is more intelligent than it seems. Adding many source files under the /src folder causes Eclipse to automatically compile all the files under that folder, unless you exclude them.
This means adding the .h to /lib folder and the .cc file to the /src folder and modifying to .cc to look for the .h file under /lib have solved the problem neatly.
To complete the compilation I had to add some flags, since the developer likes to extensively modify his library between releases.
Everything is working fine again.
I've got a project and I'm trying to make it read a .OGG file.
I've downloaded the libogg and libvorbis from here, compiled them (had some trouble figuring I had to build libogg first), then got the following files:
libogg.lib
libogg.dll
libvorbis.lib
libvorbis.dll
libvorbisfile.lib
libvorbisfile.dll
I dragged all of them in my project's Libraries folder, already added to the project, and included them in the Linker->Input (only the .lib).
Next I copied the headers to my project's Includes folder, also already added to the project, with the files:
ogg/ogg.h
ogg/os_types.h
vorbis/codec.h
vorbis/vorbisenc.h
vorbis/vorbisfile.h
Then I added some code, and I get the error "The program can't start because libvorbisfile.dll is missing from the computer."
And I'm pretty sure the file IS in the Library folder and properly defined in the properties.
Did I do something wrong along the way?
How can I figure what is wrong so I can fix it?
I already tried putting the .dll in the project's folder and in the Windows/System32 folder as well, didn't work.
The library folder is fine for your .lib files, but it's looking for the .dll at run time, which your project settings have no effect on. Windows looks in a few places for a .dll, but the easiest way to get your program to run is to put the .dlls in your working directory, which is where ever you run your executable from (probably the same directory as the .exe file).
I'm trying to start development with opencv. The problem is, until now I coul barely setup the opencv cos I cant find its header files.
I made some research regarding to this subject but none of them were realy helpful, below some of the links:
Where are the opencv2 include files?
http://answers.opencv.org/question/14712/opencvhpp-not-in-opencv2/
I made the built the files using CMAKE GUI - Built the code with MVS 11 x64 both debug and release. The compilation suceeded and I cant found libs and dlls, but no headers at all anywhere.
Checkout the pics:
Opencv2 folder
Local folder
So, now I ask, What am I missing here?
Thanks for the help
EDIT ONE:
Is that what you meant?
EDIT TWO:
I found some headers inside the source of opencv, and each of them is inside an specific folder, like core, highgui and so on - that is it? or something still wrong? cos I thought I should take the headers on the build, no on the source..
append "opencv/build/include" to your include path.
if you built the opencv libraries yourself using cmake, make sure you run the INSTALL project, and add the folder, where it installed to.
When you run make install then the include files that you need are copied to a single directory determined by the makefiles and probably displayed as a part of the output.
I know what's going on. After built the entire solution, under the "opencv" solution strcutre, there is a folder called "CMakeTargets", expand this folder, you can see "INSTALL" project, and right-click this project, then select "build" option, then after installation, all libraries and head files will be located at the correct path. Like the following picture shows:
here
Take a look at the settings that were in use when you built OpenCV - the include path must point to where the include files are for your build to have succeeded.
When I compile and run programs in Bloodshed I save everything into a a folder labeled C++ in my username folder. When I downloaded FLTK, extracted it to the C++ folder, then tried to run a program using header files from FLTK, it was unable to find the files. My guess is that when the compiler looks for the header files it's only looking in the C++ folder, and the FLTK header files are embedded in folders that are inside of the C++ folder.
I googled around for a way to somehow have file paths that include looks into when it looks for the specified header file, but I couldn't find anything. Does anyone with experience using Bloodshed know how to do this?
Most people here probably don't use DevC++, having been warned off it by people like me. DevC++ has lots of problems and is no longer being developed. You should consider
switching to Code::Blocks, which is better in just about every way.
If you have installed FLTK properly, you should now have a program called "fltk-config". That program needs to be in your PATH. You need to edit your project's settings so that the output of "fltk-config --cflags" is added to your list of compiler flags and so that the output of "fltk-config --ldflags" is added to your list of linker flags.