GBMethodData Class Reference

Overview

Describes a method or property.

Each instance or class method or property description contains at least one argument in the form of GBMethodArgument instance. Note that arguments should not be confused with method parameters, although they are very similar. In objective-c each method has a selector with optional parameter, followed by more parameters if applicable, with each parameter's selector delimited with a colon. appledoc on the other hand groups even the first part of the selector as an argument, however if the method doesn't have parameters, the argument's type and variable remain nil . This is how you can easily distinguish between different methods and properties:

• Method without parameters: methodArguments array contains a single object with [GBMethodArgument argumentName] value assigned as method name and both, [GBMethodArgument argumentTypes] and [GBMethodArgument argumentVar] set to nil . Result is optional. Example: - (void)method; .
• Method with single parameter: methodArguments array contains a single object with [GBMethodArgument argumentName] value assigned as method name, [GBMethodArgument argumentTypes] contains an array with at least one object describing the type of the parameter and [GBMethodArgument argumentVar] describing the name of the parameter variable. Result is optional. Example: - (void)method:(NSString *)var; , where argument types would contain NSString and =!= strings and argument variable var_.
• Method with multiple parameters: methodArguments array contains at least two objects, each describing it's parameter. First instance describes the base method selector name including first parameter type and variable name. Result is optional. Example - (void)method:(NSUInteger)var1 withValue:(id)var2 , where first argument would have name method , types NSUInteger and variable name var1 and second argument withValue , id and var2 .
• Properties have the same signature as methods without parameters but always have at least one result object.

To aid output templates handling, the method also prepares formatted components that can be used directly within output templates. Formatted components include all whitespace as needed to match desired coding style, so output generators can simply write the given formatted values. Although it could be argued that this should be rather part of output template, the ammount and complexity of template directives would be much greater than doing this in code. As additional bonus, we can have formatting code under unit tests to quickly verify it works as needed. And templates that really need to hande specifics, can still do so... See formattedComponents for details.