Usually I treat instance variables in Objective-c like this:
@interface MyClass : NSObject
@property (nonatomic, retain) NSMutableArray *mutableArray;
@end
@implementation MyClass
@synthesize mutableArray;
- (id) init {
if((self = [super init])) {
self.mutableArray = [NSMutableArray array];
}
return self;
}
- (void) dealloc {
[mutableArray release];
[super dealloc];
}
@end
I feel pretty comfortable w/ the above syntax. However I'm not so comfortable w/ the syntax for a 2D array instance variable like NSUInteger 2dArray[10][10]
.
What's the appropriate Objective-c syntax for a 2d array instance variable with regards to interface declaration, synthesizing getters/setters and memory management?
You don't need to allocate memory for your array; they are perfectly fine being defined in the class and they will always exist, at the same size. You therefore don't need to worry about memory management and your getter/setters should be defined manually, depending on what you want to do. For example these getter/setter methods allow getting/setting an individual value:
@interface MyClass : NSObject
{
NSUInteger _twoDeeArray[10][10];
}
- (void)setTwoDeeArrayX:(NSUInteger)x y:(NSUInteger)y value:(NSUInteger)value;
- (NSUInteger)twoDeeArrayX:(NSUInteger)x y:(NSUInteger)y;
@end
@implementation MyClass
- (void)setTwoDeeArrayX:(NSUInteger)x y:(NSUInteger)y value:(NSUInteger)value
{
_twoDeeArray[x][y] = value;
}
- (NSUInteger)twoDeeArrayX:(NSUInteger)x y:(NSUInteger)y
{
return _twoDeeArray[x][y];
}
@end
You should probably have range-checking for x and y, but you get the idea.
That's not an Objective C syntax. It's pure C syntax. You don't need to exclusively say that you want a 2D array of objc objects. Just declare/define the mutable array and add other arrays to it.
For two-demensional arrays, you can:
- Go with C arrays (like what you mentioned in the post)
- Add NSMutableArray into NSMutableArray
- Create a class to implement your version of 2D-array
If you just want to use primitive types in your array, all three are good.
For Objective-C objects, you can also go with C array with id
type but you have to manage memory allocation/deallocation yourself. 2 and 3 are better way to do this.
FYI:
- 2D arrays using NSMutableArray
- Creating a two dimensional array in Objective-C
in iOS 6 you can use subscript to define a matrix class that uses the square bracket syntax matrix[row][col]
where you can store objects and they are correctly retained by the matrix, differently than using a C array
First create a Row object, defined like this
- (id)initWithElementNumber:(NSUInteger)num {
if (self = [super init]) {
_row = [NSMutableArray arrayWithCapacity:num];
for (int j = 0; j < num; j++)
[_row addObject:[NSNull null]];
}
return self;
}
- (id)objectAtIndexedSubscript:(NSUInteger)idx {
return self.row[idx];
}
- (void)setObject:(id)object atIndexedSubscript:(NSUInteger)idx {
self.row[idx] = object;
}
@end
And then a Matrix class that uses the Row class previously defined:
@implementation UKMatrix
- (id)initWithRows:(NSUInteger)numRows columsn:(NSUInteger)numCol {
if (self = [super init])
{
_numCol = numCol;
_numRows = numRows;
_rows = [NSMutableArray arrayWithCapacity:numRows];
for (int j = 0; j < numRows; j++)
[_rows addObject:[[UKRow alloc] initWithElementNumber:numCol]];
}
return self;
}
- (id)objectAtIndexedSubscript:(NSUInteger)idx {
return self.rows[idx];
}
- (NSString *)description {
NSString *matrixDesc = @"";
for (int j = 0; j < self.numRows; j++) {
matrixDesc = [matrixDesc stringByAppendingString:@"\n"];
for (int k = 0; k < self.numCol; k++)
matrixDesc = [matrixDesc stringByAppendingFormat:@" %@ ",self[j][k]];
}
return matrixDesc;
}
@end
then you can use the Matrix with the following syntax
UKMatrix *matrix = [[UKMatrix alloc] initWithRows:4 columsn:2];
matrix[1][1] = @2;
NSLog(@"%@", matrix);