This would break key-value binding setups where a CPCollectionView's content property was hooked up to a CPArrayController's content or arrangedObjects property since the CPArrayController maintains the same object and calls setValue:forKey: with the same object the CPCollectionView already has. Therefore we cannot skip calling reloadContent even when the incoming content array is the same one we got before, because most likely the contents have changed out from under us.
CPArrayController was erroneously using canRemove to determine whether explicit removal could take place, when in fact the documentation for [NSObjectController canRemove] (NSArrayController does not override canRemove) states that it returns "YES if an object can be removed from the receiver using remove:, otherwise NO", implying that it should only be used to check whether there is a selected element that can be removed (for example, to bind the enabled state of a toolbar button to a CPArrayController's canRemove property). It should not be used to decide whether a specific object can be removed or not.
"($)path.($)variable"
"SUBQUERY(collection, $x,$x = 2)"
"sum:(arg1, ...)"
"FUNCTION(target, selector, arg1, ...)
"object[expression]"
"set UNION|INTERSECT|MINUS otherset"
Fixed a bug where CPExpression_set was evaluating to an expression instead of a set.
Replaced parsing exceptions with a generic function indicating where it failed.
When initializing a CPArrayController with a value that is not an array setContent: would make it an array by replacing value with [value]. This is consistent with how Cocoa does it, but it should first check to make sure that the content is not nil. If it is, the sensible thing to do (and what Cocoa does) is to replace it with an empty array.
This commit also includes a missing import to make CPArrayControllerTest run without error.
Added missing getters in CPExpression.
Added a type: argument to FunctionExpression initializer used by subclasses to init super with their own type.
Fixed evaluation with bindings dictionary in CPExpression_variable. Allow values in bindings to be a constant or an expression (per cocoa).
Tests: added tests for subqueryExpressions, for variableExpression evaluation where the value in the bindings dictionary can be either a constant or another expression.
Parsing a simple variable expression works (e.g. '$x = 12'), parsing a variable in a combined path (e.g.
'$x.path = 12' won't yet.
Parsing: fixed typos in built-in function names, fixed support for [FIRST], [LAST], [INDEX]
Fixed aggregates {exp1,exp2}
Fixed -expressionForFunction:selectorName:arguments:
built-in functions are now class methods
Cleaned a lot of code in built-in functions
EvaluatedObjectExpression is now a singleton
Merged inserset/minus/union Set into one file
Deleted unused expression subclasses : expression_operator and expression_assignement
Added CPCoding methods and changed key names to match Cocoa
Made KeyPathExpresion a subclass of FunctionExpression like in Cocoa
Tests
CPComparisonPredicate: fixed typo in 'between' operator evaluation, simplified 'contains' and 'in' evaluation.
Added tests in CPPredicateTest for theses fixes.
Note: this replaces the previous non Cocoa behaviour where the token field would grow height-wise with new tokens. This behaviour can still be useful, as seen in Apple Mail, but it will now require the user to subclass CPTokenField, override layoutSubview and set the frame size to the size of the scroll view's document view.
Andreas
cacaodev
Brian Donovan
Alexander Ljungberg
Klaas Pieter Annema
Antoine Mercadal
Nicholas Small
Aparajita Fishman
Daniel Stolzenberg
Blair Duncan
Randall Luecke