Saturday, 24 December 2011

Unix Domain Sockets on OSX


It's been a while since I've done anything with Unix domain sockets, and it's not the same on OSX!

If we take a look at the structure for sockaddr_un we see that the standard is:
struct sockaddr_un {
  sa_family_t sun_family; /* AF_UNIX */
  char sun_path[108]; /* pathname */
};


However, on OSX it's:
struct sockaddr_un {
  unsigned char sun_len; /* length including null */
  sa_family_t sun_family; /* AF_UNIX */
  char sun_path[104]; /* pathname */
};


Next, the standard ( or at least as far as I am aware of ) way to call bind() is as follows:

len = strlen(local.sun_path) + sizeof(local.sun_family);
if (bind(s, (struct sockaddr *)&local, len) == -1) {
  Didn't work;
}

In the above, local is the sockaddr_un struct, and the path is /var/tmp/socket.

Now as we all know ;) the normal practice is to call unlink() on the path before attempting the bind just in case the program has crashed and left the socket file lying around, however under OSX with the above code, unlink is unable to find the file.

This is because the name of the socket becomes truncated by one character, to /var/tmp/socke

It took some digging, but the correct way to call bind under OSX is as follows:

if (bind(s, (struct sockaddr *)&local, SUN_LEN(&local)) == -1) {
  perror("bind");
exit(1);


This will create the socket with the full path name, here's a complete code snippet:
#define SOCK_PATH "/var/tmp/socket"
struct sockaddr_un local;
char str[100];
int s;
if ((s = socket(AF_UNIX, SOCK_STREAM, 0)) == -1) {
 perror("socket");
 exit(1);
}

if (strlen(SOCK_PATH) >= sizeof(local.sun_path)) {
 perror("path to long!");
 exit(1);
}

local.sun_len = sizeof(local);

local.sun_family = AF_UNIX;

strcpy(local.sun_path, SOCK_PATH);

unlink(local.sun_path);

if (bind(s, (struct sockaddr *)&local, SUN_LEN(&local)) == -1) {
perror("bind");
exit(1);
}

//GO off an do what you want here

This caught me out with much head scratching for some time, so I hope it helps someone out there!


Sunday, 11 December 2011

Adobe XMP Panel - Improvements

After publishing the original post about creating a custom XMP panel ( see here ), I was contacted via the Resource Space forum about wether it was possible to create multiple sets of check boxes on the panel that related to different XMP fields, ie:

chk1,chk2,chk3 = XMP field 1
chk4,chk5,chk6 = XMP field 2

I posted on the forum that this would be easy to do using multiple handlers, and calling the relevant handler function when a checkbox is ticked.
Now many of you will note that this is not an elegant solution, but it would work, and was easy to explain!
I started thinking about this further, and came up with a more elegant way to achieve the task, that was also much easier to expand upon as there was only one place you needed to add extra code.
So this post will attempt to describe what I've done, and hopefully make it easier for others to reuse the code.

Step 1 - Layout the Panel
The first thing we need to do is layout the panel, as before this is done using the grid control, and using a naming convention for the id's of the checkboxes that relate to the label of the checkbox but without spaces.

<mx:GridItem width="100%" height="100%">
<mx:CheckBox label="Road Haulage"
id="_chkRoadHaulage"    click="updateCats(event,'docTitle')"/>
</mx:GridItem>

The label "Road Haulage" above has an id of "_chkRoadHaulage" so that we can reference it in code easily, as show in the previous article.
The major change to the layout of the checkboxes is the click event. The function that is called now has a second parameter that is the id of the XMP text input that the check box should modify.

<fi:XMPTextInput id="docTitle"
xmpPath="mc:MainCategory
xmpType="Localized"
name="docTitle"
visible="false"
/>

The second set of checkboxes have a different XMP Text Input.

<mx:GridItem width="100%" height="100%">
<mx:CheckBox label="House"
id="_chkHouse"
click="updateCats(event,'docBuilding')"/>
</mx:GridItem>

<fi:XMPTextInput id="docBuilding"
xmpPath="mc:Building
xmpType="Localized"
name="docBuilding"
visible="false"
/>

I'm also setting visible to false as I don't want to give the user the ability to manually edit the field and screw everything up ;)

Step 2 - Code Changes
Next we need to make some changes to the code.
The first thing I thought about is how we handle multiple arrays and reference them easily?
I decided the easiest way to do this was to keep the cats array and store the arrays within that refernced with a key that is the id of the XMP text input.
So the new function that gets called when the panel has completed loading is now very simple.

private function catsReadHandler(event:Event):void
{
/*  This is called once the panel has completed initialising.
        */
// create an array for each xmp field
cats['docTitle'] = new Array();
cats['docBuilding'] = new Array();
//populate the arrays and tick the right boxes
chkBoxes('docTitle');
chkBoxes('docBuilding');
}


All we have to do now is create the sub array for each XMP text input.
The next part calls a new function that reads the text input, populates the array if it contains data and then ticks the boxes that were referenced in the text input. It's basically the same as the previous catsReadHandler function, but I'm using the this[controlID] nomenclature to reference the fields and the checkboxes.
This gets called for each set of checkboxes / XMP field in the panel.

Here's the function:

private function chkBoxes(xmpFieldName: String):void
{
// this takes the name of the xmp field, populates the array, and ticks the checkboxes
var outStr:String;
var catsString: String = this[xmpFieldName].text;
var chkName:String
var catStr:String;
if (catsString.length > 0)
{
// split the text by commas and populate the array
cats[xmpFieldName] = catsString.split(",");
for (var i:int = 0; i < cats[xmpFieldName].length; ++i) 
{
// remove spaces from the cat name and add to _chk to ref the box
chkName = "_chk";
catStr = cats[xmpFieldName][i];
catStr.replace(/s+/g, "");
chkName += catStr;
try {
// tick the box!
this[chkName].selected=true;
} catch (someError:Error)
{
this[xmpFieldName].text=i + " : " + " " + chkName + " " + someError.message;
}
}
}
}

As you can see, it's pretty much the same as before, but putting this in a separate function enables us to simplify the catsReadHandler function and only have to add extra code there.

The final piece of the puzzle is the function that gets called when a user checks / un-checks a box.
This now takes the id of the XMP text input to update, and so we can reference that directly without having to write a special function for each set of check boxes & XMP fields.

private function updateCats(event: Event,fieldName: String):void
{
//we use event.target to identify the control that triggered the function :)
var catName:String = event.target.label;
var outStr:String;
var cName:String;
var delItem:Boolean;
var localArray:Array;
// copy the main array so we can manipulate it.
localArray = cats[fieldName];
if (localArray.length == 0)
{
localArray.push(catName);
} else {
for (var i:int = 0; i < localArray.length; ++i) {
if (localArray[i] == catName)
{
localArray.splice(i,1);
delItem = true;
}
}
if (!delItem)
{
localArray.push(catName);
}
}
cats[fieldName]=localArray;
outStr = cats[fieldName].toString();
this[fieldName].text = outStr;
this[fieldName].dispatchEvent(new Event(Event.CHANGE));
} 

One of the major changes here apart from calling the function with the field id is that I'm copying the array before manipulating it and then copying it back to the main array after the manipulation is done.
I know in theory I shouldn't need to do this, but I was having issues manipulating the main array directly, so I chose this method because I couldn't be bothered to debug it any further than I already had ;)

The complete file can be found here, feel free to play around with it.




Thursday, 8 December 2011