Jay Tuley

Recently Elsewhere

Cocoa collection subclassing | Jens/Log disagrees Faux Collection Class Subclassing | AgentM and both have some bits on class clusters. AgentM’s is a rant about not being able to easily subclass collections, which I am sympathetic too recalling the joy of programing in smalltalk with its rich collection hierarchy. Jen’s is basically a response saying Class Clusters can be frustrating but their okay, just drink the koolade AgemtM. If you explain Class Clusters to someone, they makes sense, but when you try to use Apple’s or don’t know they are there, frustrating things can happen, and I don’t think that it’s really the class clusters fault. Making a composite class by wrapping or forwarding messages isn’t that much of a hassle, subclassing anything requires some knowledge of the class and Apple does document approaches to subclass class clusters, so I’m on the koolade camp on AgentM’s rant. However, I believe there are some bugs in the implementation of a few of Apple’s class clusters that cause many to think that all Class Clusters are evil.

FUD

I think there’s a lot of fear, uncertainty, and doubt in regards to class clusters. There really shouldn’t be though, Apple has a nice page describing the benefits of Class Clusters — Cocoa Objects: Class Clusters

See a class cluster really is really just a sneaky use of a Factory Method (GOF 107) to instantiate a subclass from an abstract superclass. It’s sneaky because its hides it all. alloc, rather than actually allocating, returns a Creator (probably a singleton) and init actually allocates your real product and returns it. As demonstrated here:

int main (int argc, const char * argv[]) {
    NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init];
    id obj =[NSArray alloc];
    NSLog(NSStringFromClass([obj class]));
    NSLog(NSStringFromClass([[obj init] class]));
    [pool release];
    return 0;
}

2006-03-17 22:55:15.100 TestDataTypes[6105] NSPlaceholderArray
2006-03-17 22:55:15.101 TestDataTypes[6105] NSCFArray

This is all good, because the classes returned are all subclasses and being the good little object oriented programers we are, we never check what a class actually is. Instead, when need to runtime type check it’s only to determine if it’s a subclass or if it responds to a method or protocol and that should be just fine with class clusters.

If only using standard best practices when runtime type checking would keep us safe, unfortunately a few of Apple’s class clusters, mainly collection classes, are implemented in such a way that they break expected inheritance and expected polymorphism by advertising the wrong public interface for the returned private subclass (I don’t know if this was the case for NeXT or not, but I suspect it wasn’t). I’ll continue with some examples of where this implementation goes wrong.

Dumb Responses from Class Clusters with Multiple Public Superclasses

Here’s a Pop Quiz: What should this statement return?

[[NSArray array] isKindOfClass:[NSMutableArray class]];

Ideally it should return false because when we see the documentation on these public abstract classes, NSMutableArray inherits from NSArray.

However, actually this statement returns true because the private subclasses returned for both public abstract superclasses are the same class inherited from NSMutableArray known as NSCFArray.

Okay so the purpose of having the same class is for easy toll free bridging with Core Foundation, but you know, as long as the instance variables are the same, is Core Foundation really going to care about new methods or a different hierarchy in obj-c land? I’m thinking no in theory, but I’ve not tested and I’m much happier not digging into the C-ness of Obj-C and someone can prove me wrong or right if they wish. However, if you really really wanted to keep the same private class type between the two interfaces, you can always fake the expected hierarchy by making isKindOfClass: call isSubclassOfClass: on the appropriate abstract superclass’s metaclass, and really no one would be the wiser, Core Foundation or otherwise.

That was problem #1, and so you might be thinking to yourself well I don’t test for inheritance with cocoa programing often, most of the time I’ll test for specific methods out of habit, because cocoa programers tend to use informal protocols rather than having deep hierarchies anyway, so it’s not that big of a deal.

So our second pop quiz: What is the result of this statement?

[[NSArray array] respondsToSelector:@selector(addObject:)];

NSArray is not supposed to respond to addObject:. I know that, you know that, but a program is generally going to believe respondsToSelector: over what we think, and respondsToSelector: says true. Of course if you run this line:

[[NSArray array] addObject:@“test”];

you get an exception. Think about when you have a bunch of objects in a collection with various types, some may respond to addObject:, some don’t, and you want to use runtime checking to manipulate them appropriately. How are you suppose to take advantage of polymorphism of this method in this case without being able to trust respondsToSelector:.

There are simple work arounds to these problems, because they are only a handful of classes, but this is clearly a poor implementation of a class cluster, and I haven’t tested every class but I think all Mutable-NonMutable class hierarchies in Foundation on OS X suffer from this bug (Although GNUStep is A-OK from looking at the source), and I don’t think there’s a reason not to fix it, being that they are class clusters, all apple needs to modify are the private subclasses and people aren’t going to be using those methods currently because they return the wrong values that aren’t useful when wrong.

Demonstration of Problem and Fixability

I did write a few Unit Tests to clearly demonstrate this problem with NSArray. It also has a second target called ProperClassCluster that overrides some NSCFArray methods (using categories) to fix responses in inheritance & method testing. In that target I also swizzled the addObject: method to provide an example of returning the correct exception when calling a method that doesn’t exist on an NSArray. A real fix wouldn’t require these hacks, but this works for demonstration purposes. Download ProperClassCluster.zip (20k)

Continue reading "Class Clusters"

So I started making some default scripts and other script examples for NicePlayer. I then realized that I needed keyboard shortcuts on the CocoaScriptMenu. DVD Player’s script menu just uses numbers in order, I don’t like that, another solution I considered was to have some kind of separate config, either gui or text file. However, since renaming is required to order and change the menu text in CocoaScriptMenu, that could get a little complicated.

Keyboard Shortcuts

The solution I settled on was to add the keyboard shortcut itself to the filename. I feel this implementation ended up working really well. The main worry I had was that user error could end up disrupting the host application’s keyboard shortcuts, but the script menu is loaded later than the other menus, so it’s menu shortcuts have the least precedence and thus show up blank if a user tries to duplicate a shortcut. To facilitate adding keyboard combos to file name, I came up with an ASCII representation for the modifiers that kinda look like the real symbols and all require shift to create them (so they won’t reduce the possible representable keyboard short cuts). The symbols are as follows

∗ – Command

$ – Shift

^ – Control

% – Option

To create a short cut you add it between two curly braces in the file name before the file extension, ordering the modifiers before the key character, such as Hello World {∗^$H}.scpt for command-control-shift-H.

However after implementing this, I realized most people are going to be using a filesystem that supports unicode, and this can look a lot prettier. So I added support for these shortcuts, not just in the ascii, but using this UTF8:

⌘ – Command (0×2318 PLACE OF INTEREST SIGN)

⇧ – Shift (0×21E7 UPWARDS WHITE ARROW)

⌃ – Control (0×2303 UP ARROWHEAD)

⌥ – Option (0×2325 OPTION KEY)

And you can surround them by LEFT & RIGHT SQUARE BRACKET WITH QUILL (0×2045 & 0×2046) instead of curly brackets. Such as Hello World ⁅⌃⇧⌘H⁆.scpt

While UTF8 looks a bit neater, for safety sake, all my default NicePlayer scripts will use the ASCII way. And the ASCII way is also easier to type, so I’m guessing it will end up being the preferred way, but it’s left up to the user.

You can even add Function Key shortcuts by just typing out {F13} or {⌃⇧F2} or even {∗F16}.

and this scheme should cover a great majority of the possible keys, getting all of them however would require more parsing, which I don’t really see the need at this point, but is possible in the future.

The down side is that I added a new instance variable to the Command classes, so if you do any subclassing of these classes you’ll need to recompile those subclasses with this framework Versioning and Compatibility (not a big deal as CocoaScriptMenu is meant to be embedded not shared).

Count of Scripts

I added a method to the menu generator to count the number of scripts, this can be used to determine whether or not scripts are currently installed.

Panther Compatibility

I did a little bit of hacking to give developers basic usage of the framework under 10.3.9, but I haven’t done any testing with other 10.3ness such as compiling under GCC 3 or actually compiling on 10.3. That said with these compatibility additions there’s a little more work needed when adding support for new filetypes or being more specific with filetypes, as I pretty much just fake the UTI stuff when it runs in Panther and thus more faking or a better solution is required, but if you don’t care about 10.3 then you can continue to not worry about it.

PS

MarsEdit rocks, I wrote this post in it and didn’t have to worry about the UTF8 characters, they were safely converted to html Entities.

Continue reading "Cocoa Script Menu Revised 1.01"

The Sunday Project

So last sunday I started on something new, really it’s a feature for NicePlayer, but also a feature in a lot of other existing apps out there, and could be useful in a lot of cocoa programs that don’t have this feature yet, so I wrote my implementation as an embedable framework, and am releasing under the MPL/LGPL/GPL (my latest preferred OSI approved license for those who notice what license I release under).

So here’s a riddle, what do iTunes, DVD Player, Xcode, FlySketch, NetNewsWire, MarsEdit and BBEdit all have in common?

Give up?

They all have one of these (more or less):

Their own script menu.

Uses in NicePlayer

Having a script menu in NicePlayer has been something in the back of my mind for a while. There are some features requests, while really simple, are very specific to individual user needs and we can’t justify adding a feature. Sometimes there are features requests that just don’t fit into Robert’s or my idea of NicePlayer, and we barely have enough time to add the features we want to add, so in the next release users can add their own menu commands in this script menu.

One of the features introduced in 0.92 of NicePlayer was an option to remove the fixed aspect ratio. Maybe you want to distort the movie, maybe the aspect ratio is just slightly off, this feature works in those cases, and only adds one more menu to the window and is inline with adding basic window options in the Window menu that we did before. However this isn’t useful when someone has a lot of media that is consistently using non-square pixels. In the case of standard media formats, the proper behavior of NicePlayer should be to automatically adjust (this feature has been added for the DV codec when using the CoreVideo plugin in the next version of NicePlayer 0.93). However there are people out there for some reason, how have media that is encoded with non-square pixels for no standard reason, and there isn’t a way to detect the correct aspect ratio. It turns out writing an AppleScript with the current NicePlayer dictionary to set a Window to a different aspect ratio is pretty trivial, so this is one of the scripts we’ll likely include in the next release (since its one the scripts i’ve been using to test anyway).

I think the main benefit of having the script menu, will be nicer integration with other apps. Whether having it integrate with a cataloging app, Toast, or just organizing with the finder, there seem to be many potential uses in this respect.

CocoaScriptMenu.Framework in Your Own Program

The framework is called CocoaScriptMenu.Framework and available on my software page. It’s not my favorite name of the software I’ve written, but I think it’ll help in being google-able for those wanting it’s feature. I’m releasing it as version 1.0, and as I said before under the MPL/LGPL/GPL license. The basic way to use it is to:

1. Add the framework binary to your project
2. Add the framework to a copy phase that puts it into the Frameworks folder of your app bundle
3. Add this line [[CSMScriptMenu sharedMenuGenerator] updateScriptMenu]; somewhere right after the nib loads (while including <CocoaScriptMenu/CocoaScriptMenu.h> for that file of course)

And that will give you a typical script menu once you compile and run your app.

Features

I wrote typical, as not all script menus are the same. I tried to add what I felt were the best features of all the script menus, while being able to behave, depending on how you use it, like 80% of the script menus I’ve seen with out any extra customization. Some of the features are

• It looks for scripts in Application Support/AppName/Scripts in all domains.
• You can make submenus by nesting folders.
• You can add a menu separator by adding a file or folder with a dash as it’s name.
• You can order scripts, separators, and folders using a two digit prefix on the file name (two digit prefix not required).
• It runs AppleScripts, Automator Files (.workflow), Application Bundles, and Shell Scripts (sh/python/perl/whatever but remember you need executable permission for shell scripts).
• It automatically updates the menu, without relaunching the application, as you add items or folders nested inside the script folder, unless that folder didn’t exist at Application Launch.

Extending w/o Modifying

The singleton [CSMScriptMenu sharedMenuGenerator] has 4 optional delegate methods that allow you to keep the core functionality but make some slightly different script menus without having to modify the source (although modification is certainly an option. (warning most of this is untested as I use only the default implementation of each of these in NicePlayer)

-(NSMenuItem*)showScriptFolderMenuItem;

If your delegate implements this method, it needs to return a NSMenuItem that should be used instead of the Open Scripts Folder menu item. An example would be a menuitem with a submenu with an open command for each each script folder in the domains. The default behavior is a menuitem to open the first location returned by the delegate method -(NSArray*)scriptLocations; in the finder.

-(NSArray*)argumentsForShellScripts;
The default implementation returns nil. Implement this method in your delegate if you want to pass string arguments to any shell script. (is not used for AppleScripts, Application Bundles or Workflows)

-(id)scriptMenuItemOrItems;
This should return something that has a subMenu or menu getter for an NSMenu. This NSMenu is what the script menu items get added too. You may also return an array to add the script menu in multiple places in your app. The default implementation adds a Script Menu in-front of the Help menu and returns it.

-(NSArray*)scriptLocations;
The default implementation, returns the paths to Application Support/MainBundleName/Scripts for every domain that the path exists, with the User domain guaranteed first. Also creates the folder for the user domain if it doesn’t exist.

Extending by Modifying

So in version 1.0 the script running implementations are very basic. They are setup as a Class Cluster, with the CSMCommand class providing the public interface and several subclasses that implement script running for various types of scripts or executables.

Diagram of class hierarchy:

alloc on CSMCommand returns a singleton instance of CSMPlaceholderCommand. CSMPlaceholderCommand’s initWithScriptPath: works as parameterized factory method and depending on the path passed in returns the correct concrete implementation allocated and initialized.

So to extend the implementation of script execution for an existing file type, you would just modify one of the subclasses. To add a new filetype you would modify the initWithScriptPath: factory method and add a new subclass.

initWithScriptPath: uses Apple’s Uniform Type Identifiers to determine which concrete class to instantiate, so order does matter, make sure you add the more specific type checks in the beginning of the method and the more general towards the end.

There are a lot of ways that I can think of that the concrete script execution classes could be improved, however for NicePlayer these all work well, thus I figured for version 1.0 it was better to stick with the basic implementation and release it now, rather than try and over engineer. So I’ll wait and see if people need more or not, not to mention they have the option of contributing code.

Updated at Cocoa Script Menu Revised 1.01

Continue reading "My Sunday Project - Reusable Cocoa Script Menu"