I'm developing an extension of eclipse CDT plug-in which has to parse the c++ code and find all
enumerations (names)
enumerators (names)
enumerator values (numbers)
Suppose .cpp file contains following text
enum SomeEnum
{
One = 0,
Two = 1,
Three = 2,
Four = 3,
maxNum
};
The pluging should output following :
Enumeration - SomeEnum
Enumerators - One, Two, Three, Four, maxNum
Values - 0, 1, 2, 3,
The visitor method of ASTVisitor inherited class finds the enumerations this way
public int visit(IASTDeclaration declaration) {
if (declaration instanceof IASTSimpleDeclaration) {
IASTDeclSpecifier specifier = ((IASTSimpleDeclaration)declaration).getDeclSpecifier();
if (specifier instanceof IASTEnumerationSpecifier) {
IASTEnumerationSpecifier enumSpecifier = (IASTEnumerationSpecifier)specifier;
IBinding binding = enumSpecifier.getName().resolveBinding();
System.out.println(enumSpecifier.getName());
}
}
return PROCESS_CONTINUE;
}
Question : How can I iterate over the enumerators and their corresponding values of found enumeration ?
Well figured out on my own
public int visit(IASTDeclaration declaration) {
if (declaration instanceof IASTSimpleDeclaration) {
IASTDeclSpecifier specifier = ((IASTSimpleDeclaration)declaration).getDeclSpecifier();
if (specifier instanceof IASTEnumerationSpecifier) {
IASTEnumerationSpecifier enumSpecifier = (IASTEnumerationSpecifier)specifier;
// Get the current enumeration name
String enumerationName = enumSpecifier.getName().toString();
IASTEnumerator[] enumerators = enumSpecifier.getEnumerators();
for (IASTEnumerator enumerator : enumerators)
{
System.out.println(enumerator.getName());
if (enumerator.getValue() instanceof IASTLiteralExpression ) {
System.out.println(enumerator.getValue());
}
}
}
}
return PROCESS_CONTINUE;
}
Related
In Swift you can pass a type into a function like so:
func foo<T>(_ type: T.Type)
foo(String.self)
Is there any equivalent in Kotlin?
I don't want to specify the type in the generic arguments, because there are multiple generic arguments and it would be annoying to have to specify them all when the compiler can work it out itself except for this one. Also I don't want to use KClass as that requires the type to by Any which I don't want to make a requirement for the type.
you can you class.java.typeName
/**
* Return an informative string for the name of this type.
*
* #return an informative string for the name of this type
* #since 1.8
*/
public String getTypeName() {
if (isArray()) {
try {
Class<?> cl = this;
int dimensions = 0;
do {
dimensions++;
cl = cl.getComponentType();
} while (cl.isArray());
StringBuilder sb = new StringBuilder();
sb.append(cl.getName());
for (int i = 0; i < dimensions; i++) {
sb.append("[]");
}
return sb.toString();
} catch (Throwable e) { /*FALLTHRU*/ }
}
return getName();
}
foo(Int::class.java.typeName)
foo(String::class.java.typeName)
Before you vote on this question I would like to how dumb what I am trying to do is. Maybe I still don't understand properly enums.
So, I am working on a project that uses an Obj-C framework. This framework contains enums:
typedef enum : NSInteger
{
kImageSizeUnknown = 0,
kImageSize75,
kImageSize110,
kImageSize170,
kImageSize220,
kImageSize300,
kImageSize450,
kImageSize720,
kImageSize1080,
/* Size aliases */
kImageSizeThumbnail = kImageSize75,
kImageSizeSmall = kImageSize170,
kImageSizeMedium = kImageSize450,
kImageSizeLarge = kImageSize720,
kImageSizeXLarge = kImageSize1080
} GnImageSize;
I want somehow be able to declare a Swift enum that returns values of the Obj-C enum (That might be the silly part).
That is how I have at the moment.
enum GNImageSize:Int, CaseIterable{
case thumbnail
case sizeSmall
func toGnImageSize() -> GnImageSize {
switch self {
case .thumbnail:
return kImageSizeThumbnail
case .sizeSmall:
return kImageSizeSmall
}
}
static func toGnImageSize(sizeType:GNImageSize) -> GnImageSize {
switch sizeType {
case .thumbnail:
return kImageSizeThumbnail
case .sizeSmall:
return kImageSizeSmall
}
}
}
However when I do:
enum GNImageSize:Int, CaseIterable{
case thumbnail = GnImageSize.kImageSizeThumbnail
case sizeSmall = GnImageSize.kImageSizeSmall
}
I get the following error message:
Raw value for enum case must be a literal
Thank you.
EDIT
Function that comunicates with Obj-c functions
func getArtworkURL(forImageType imageType:GNImageSize, shouldFindAlternatives:Bool, highQualityFirst:Bool)->URL?{
if let asset = coverArt()?.asset(GnImageSize(rawValue: imageType.rawValue)), let assetURL = asset.urlHttp(){
return URL(string:assetURL)
}
else{
if shouldFindAlternatives{
if highQualityFirst{
for size in GNImageSize.allCases.reversed(){
if let asset = coverArt()?.asset(GnImageSize(rawValue: size.rawValue)), let assetURL = asset.urlHttp(){
return URL(string:assetURL)
}
}
}
else{
for size in GNImageSize.allCases{
if let asset = coverArt()?.asset(GnImageSize(rawValue: size.rawValue)), let assetURL = asset.urlHttp(){
return URL(string:assetURL)
}
}
}
}
}
return nil
}
Where
-(nullable GnAsset*) asset: (GnImageSize)imageSize;
The five size aliases have the (raw) values 1, 3, 6, 7, 8 so declare a Swift enum
enum GNImageSize : Int {
case thumbnail = 1
case small = 3
case medium = 6
case large = 7
case xLarge = 8
}
To use the Int value in Swift use for example
GNImageSize.thumbnail.rawValue
Alternatively create a custom enum with static properties to map the types
enum GNImageSize {
static let thumbnail = GnImageSize(0)
static let small = GnImageSize(3)
static let medium = GnImageSize(6)
static let large = GnImageSize(7)
static let xLarge = GnImageSize(8)
}
I don't understand that in 2018 ObjC frameworks still use the Stone-age syntax typedef enum : NSInteger { ... } Foo; rather than Swift compliant syntax typedef NS_ENUM (NSInteger, Foo) { ... }; The latter syntax exists for 6 years (iOS 6, macOS 10.8).
That raw-value style enum grammar error.
raw-value-assignment → = raw-value-literal
raw-value-literal → numeric-literal | static-string-literal | boolean-literal
So, only numeric(numbers like -7, 0x10, 0b010), static string(characters in quotes, like "foo") and boolean(true or false) literals are allowed there.
Anything else won't work.
Previously I was using this code:
private val mItems = ArrayList<Int>()
(1..item_count).mapTo(mItems) { it }
/*
mItems will be: "1, 2, 3, 4, 5, ..., item_count"
*/
Now, I am using a class instead of Int, but the class has Int member with name id.
class ModelClass(var id: Int = 0, var status: String = "smth")
So how can I use this method to fill the ArrayList in similar way?
//?
private val mItems = ArrayList<ModelClass>()
(1..item_count).mapTo(mItems) { mItems[position].id = it } // Something like this
//?
From the mapTo documentation:
Applies the given transform function to each element of the original collection and appends the results to the given destination.
Therefore, you just need to return the elements you want:
(1..item_count).mapTo(mItems) { ModelClass(it) }
If you are OK with any MutableList (which is often ArrayList or similar):
val mItems1 = MutableList(item_count) { i -> i }
val mItems2 = MutableList(item_count) { ModelClass(it) }
I cannot seem to find anywhere, any examples on how to make use of the GroupFormatter delegate to allow me to add footers to my groups when using the ObjectListView control.
Does anyone have any examples that could demonstrate this? I want to remove the text from the group header and add a footer (different text per footer). As well as changing font, etc.
Any examples would be very helpful.
You can analyze the code for the
public void MakeGroupies<T>(T[] values, string[] descriptions, object[] images, string[] subtitles, string[] tasks)
method of the ObjectListView class. That explicitly sets the GroupKeyGetter, GroupKeyToTitleConverter and GroupFormatter property delegates.
This is C# but your VB adaptation should be straightforward. I am using this small test class as the object type to bind to the list view.
public class TestClass
{
private readonly string _s;
private readonly float _f;
public TestClass( string p1, float p2 )
{
this._s = p1;
this._f = p2;
}
[OLVColumn(DisplayIndex = 1, Name="S", Title="String")]
public string S {get {return this._s;}}
[OLVColumn( DisplayIndex = 2, Name = "F", Title = "Float" )]
public float F {get {return this._f;}}
}
So as not to manually define column traits I am using attributes inside the bound object and a
BrightIdeasSoftware.Generator.GenerateColumns( this.olv, typeof( TestClass ) );
call in the form/user control where I am using the list view. In fact here is the method that completely isolates ObjectListView configuration:
void SetData( TestClass[] objects )
{
// build list columns
Generator.GenerateColumns( this.olv, typeof( TestClass ) );
// use groups and make current column the priimary sort column
this.olv.ShowGroups = true;
this.olv.SortGroupItemsByPrimaryColumn = false;
// loop through columns and set properties
foreach( OLVColumn col in this.olv.Columns )
{
col.Groupable = true;
col.Sortable = true;
if( col.Name == "F" )
{
col.MakeGroupies<float>( new float[] { 10f, 100f, 1000f }, new string[] { "<10", "10-100", "100-1000", ">1000" } );
}
else if( col.Name == "S" )
{
col.UseInitialLetterForGroup = false;
//
col.GroupKeyGetter = ( obj ) =>
{
TestClass tc = (TestClass)obj;
switch( char.ToLower( tc.S[0] ) )
{
case 'a':
case 'e':
case 'i':
case 'o':
case 'u': return true;
default: return false;
}
};
//
col.GroupKeyToTitleConverter = ( o ) => { bool b = (bool)o; return b ? "vowel" : "consonant"; };
//
col.GroupFormatter = ( /*OLVGroup*/ group, /*GroupingParameters*/ parms ) =>
{
string s = string.Format ("{0} {1}", group.GroupId, group.Id);
//group.BottomDescription = "BottomDescription: " + s;
//group.TopDescription = "TopDescription: " + s;
group.Footer = "Footer: " + s;
};
}
}
//
this.olv.RebuildColumns();
//
this.olv.SetObjects( objects );
}
You will definitely have one different footer per each group.
I am not sure how to write comparator for Collectionos.binarySearch(). Can anyone help ? sample code:
List<Object> list1 = new ArrayList<>();
List<List<Object>> list2 = new ArrayList<>();
//loop starts
// adds elements into list1
list1.add(values);//values is an object containing elements like [3, John, Smith]
if (list2.size() == 0) {
list2.add(list1);//first element
} else {
if (index >= 0) {
int index = Collections.binarySearch(list2, list1, comparator);
list2.add(index, list1);//I want to add these elements in ascending order ?
}
}
//loop ends
How do I write comparator, so that elements in list 2 are added in ascending or descending order.
You can use an anonymous class which implements a Comparator<List<Object>>:
int index = Collections.binarySearch(list2, list1, new Comparator<List<Object>>() {
#Override
public int compare(List<Object> o1, List<Object> o2) {
// Your implementation here
return 0;
}
});
You could implement an IComparer<List<Object>> class, or use a lambda expression.
You just need to create a class that implements the Comparator interface.
For example, you can do this inline with an anonymous class:
Comparator<List<Object>> comparator = new Comparator<List<Object>>() {
#Override
public int compare(List<Object> x, List<Object> y) {
// custom logic to compare x and y here. Return a negative number
// if x < y, a positive number if x > y, and 0 otherwise
}
};
Collections.binarySearch(list, comparator);