File size: 3,744 Bytes
d4035c1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 |
<!DOCTYPE group PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<group>
<p>These instructions show how to setup a basic <b>VLFeat</b> project
with Apple Xcode. For the sake of simplicty, we create a command
line tool written in C. However, these steps apply with minor
modifications to other project types and to the C++ lanuage.</p>
<p>First, let us create a new project
called <code>vlfeat-client</code>. Open Xcode and select <b>File
> New Project > Command Line Utility > Standard Tool</b>
and click <b>Choose</b>. Give a name to your project (in our
case <code>vlfeat-client</code>), and click <b>Save</b>.</p>
<div class="figure">
<img alt="Xcode new project" src="%pathto:root;images/using-xcode-new.png"/>
</div>
<p>Now we need to add <b>VLFeat</b> to the C compiler include search
path. To do this, select the <code>vlfeat-client</code> target and
open the information panel (the blue button,
or <b>Command-i</b>). Then select the <b>Build</b> panel, search for
the field <b>Header Search Paths</b>, and add
<b>VLFeat</b> root path (in our case this is
just <code>~/src/vlfeat</code>).</p>
<img alt="Xcode info" src="%pathto:root;images/using-xcode-info.png"/>
<p>Next, we add the <code>libvl.dylib</code> library file to the
project resources so that Xcode links against it. To do this, drag
and drop the <code>libvl.dylib</code> file (in our example
<code>~/src/vlfeat/bin/maci/libvl.dylib</code>) to the left panel and click
<b>Add</b>.</p>
<img alt="Xcode dylib" src="%pathto:root;images/using-xcode-dylib.png"/>
<p>Next, edit the <code>main.c</code> source file and type the following code:</p>
<pre>
#include <vl/generic.h>
int main (int argc, const char * argv[]) {
VL_PRINT ("Hello world!") ;
return 0;
}
</pre>
<img alt="Xcode edit" src="%pathto:root;images/using-xcode-edit.png"/>
<p>If you try to build the project, it should compile without errors
(if you are using C++, do not forget to wrap the <code>include</code>
statements in a <code>extern "C" {}</code> block). However, if you try
to run the program, it will fail, complaining that it cannot find the
library image.</p>
<img alt="Xcode error" src="%pathto:root;images/using-xcode-err.png"/>
<p>The reason is that <code>libvl.dylib</code> is compiled with the
library <code>install_name</code> equal
to <code>@loader_path/libvl.dylib</code>. This causes the run-time
loader to look for the library in the same directory of the
executable. There are two ways around this problem: The first is to
install the library in a standard location
(e.g. <code>/usr/local/lib</code>) and use the <code>otool</code>
command to change the
library <code>install_name</code>. The other is to simply copy
the <code>libvl.dylib</code> file in the executable directory. Here we
demonstrate the second technique.</p>
<p>To copy <code>libvl.dylib</code> in the executable directory, we
add a <b>Copy Files</b> build phase to the project. Right-click
the <code>vlfeat-client</code> target in the project panel and select
<b>Add > New Build Phase > New Copy Files Build
Phase</b>. Select <b>Destination: Executables</b>. Then drag-and-drop
the <code>libvl.dylib</code> item from the panel to the <b>Copy
Files</b> build phase.</p>
<img alt="Xcode copy" src="%pathto:root;images/using-xcode-copy.png"/>
<img alt="Xcode copy" src="%pathto:root;images/using-xcode-copy-2.png"/>
<p>Now rebuild the project, and run it. It should run correctly,
and if you open the debugger console you should see this:</p>
<img alt="Xcode ok" src="%pathto:root;images/using-xcode-ok.png"/>
</group>
|