…dArray` class (apache#37613)

### Rationale for this change

Currently, there is no way to easily convert an `arrow.array.ChunkedArray` into a corresponding MATLAB array, other than (1) manually iterating chunk by chunk, (2) calling `toMATLAB` on each chunk, and then (3) concatenating all of the converted chunks together into one contiguous MATLAB array.

It would be helpful to add a toMATLAB method to `arrow.array.ChunkedArray` that abstracts away all of these steps.

### What changes are included in this PR?

1. Added `toMATLAB` method to `arrow.array.ChunkedArray` class
2. Added `preallocateMATLABArray` abstract method to `arrow.type.Type` class. This method is used by the `ChunkedArray` `toMATLAB` to pre-allocate a MATLAB array of the expected class type and shape. This is necessary to ensure `toMATLAB` returns the correct MATLAB array when the `ChunkedArray` has zero chunks. If `toMATLAB` stored the result of calling `toMATLAB` on each chunk in a `cell` array before concatenating the values, `toMATLAB` would return a 0x0 `double` array for zero-chunked arrays. The pre-allocation approach avoids this issue.
3. Implement `preallocateMATLABArray` on all `arrow.type.Type` classes.
4. Added an abstract class `arrow.type.NumericType` that all classes representing numeric data types inherit from. `NumericType` implements `preallocateMATLABArray` for its subclasses.

### Are these changes tested?

Yes. Added unit tests to `tChunkedArray.m`.

### Are there any user-facing changes?

Yes. Users can now call `toMATLAB` on `ChunkedArray`s.

**Example**

```matlab

>> a = arrow.array([1 2 NaN 4 5]);
>> b = arrow.array([6 7 8 9 NaN 11]);
>> c = arrow.array.ChunkedArray.fromArrays(a, b);
>> data = toMATLAB(c)

data =

     1
     2
   NaN
     4
     5
     6
     7
     8
     9
   NaN
    11

```

* Closes: apache#37597

Authored-by: Sarah Gilmore <[email protected]>
Signed-off-by: Kevin Gurney <[email protected]>

…dArray` class (apache#37613)

### Rationale for this change

Currently, there is no way to easily convert an `arrow.array.ChunkedArray` into a corresponding MATLAB array, other than (1) manually iterating chunk by chunk, (2) calling `toMATLAB` on each chunk, and then (3) concatenating all of the converted chunks together into one contiguous MATLAB array.

It would be helpful to add a toMATLAB method to `arrow.array.ChunkedArray` that abstracts away all of these steps.

### What changes are included in this PR?

1. Added `toMATLAB` method to `arrow.array.ChunkedArray` class
2. Added `preallocateMATLABArray` abstract method to `arrow.type.Type` class. This method is used by the `ChunkedArray` `toMATLAB` to pre-allocate a MATLAB array of the expected class type and shape. This is necessary to ensure `toMATLAB` returns the correct MATLAB array when the `ChunkedArray` has zero chunks. If `toMATLAB` stored the result of calling `toMATLAB` on each chunk in a `cell` array before concatenating the values, `toMATLAB` would return a 0x0 `double` array for zero-chunked arrays. The pre-allocation approach avoids this issue.
3. Implement `preallocateMATLABArray` on all `arrow.type.Type` classes.
4. Added an abstract class `arrow.type.NumericType` that all classes representing numeric data types inherit from. `NumericType` implements `preallocateMATLABArray` for its subclasses.

### Are these changes tested?

Yes. Added unit tests to `tChunkedArray.m`.

### Are there any user-facing changes?

Yes. Users can now call `toMATLAB` on `ChunkedArray`s.

**Example**

```matlab

>> a = arrow.array([1 2 NaN 4 5]);
>> b = arrow.array([6 7 8 9 NaN 11]);
>> c = arrow.array.ChunkedArray.fromArrays(a, b);
>> data = toMATLAB(c)

data =

     1
     2
   NaN
     4
     5
     6
     7
     8
     9
   NaN
    11

```

* Closes: apache#37597

Authored-by: Sarah Gilmore <[email protected]>
Signed-off-by: Kevin Gurney <[email protected]>