Package org.codehaus.groovy.ast.tools

Class WideningCategories

java.lang.Object
org.codehaus.groovy.ast.tools.WideningCategories
public class WideningCategories
extends Object
This class provides helper methods to determine the type from a widening operation for example for a plus operation.

To determine the resulting type of for example a=exp1+exp2 we look at the conditions isIntCategory(ClassNode), isLongCategory(ClassNode), isBigIntCategory(ClassNode), isDoubleCategory(ClassNode) and isBigDecCategory(ClassNode) in that order. The first case applying to exp1 and exp2 is defining the result type of the expression.

If for example you look at x = 1 + 2l we have the first category applying to the number 1 being int, since the 1 is an int. The 2l is a long, therefore the int category will not apply and the result type can't be int. The next category in the list is long, and since both apply to long, the result type is a long.

  • Constructor Details

    • WideningCategories

      public WideningCategories()

  • Method Details

    • isInt

      public static boolean isInt​(ClassNode type)
      Used to check if a type is an int or Integer.
      Parameters:
      type - the type to check

    • isDouble

      public static boolean isDouble​(ClassNode type)
      Used to check if a type is an double or Double.
      Parameters:
      type - the type to check

    • isFloat

      public static boolean isFloat​(ClassNode type)
      Used to check if a type is a float or Float.
      Parameters:
      type - the type to check

    • isIntCategory

      public static boolean isIntCategory​(ClassNode type)
      It is of an int category, if the provided type is a byte, char, short, int.

    • isLongCategory

      public static boolean isLongCategory​(ClassNode type)
      It is of a long category, if the provided type is a long, its wrapper or if it is a long category.

    • isBigIntCategory

      public static boolean isBigIntCategory​(ClassNode type)
      It is of a BigInteger category, if the provided type is a long category or a BigInteger.

    • isBigDecCategory

      public static boolean isBigDecCategory​(ClassNode type)
      It is of a BigDecimal category, if the provided type is a BigInteger category or a BigDecimal.

    • isDoubleCategory

      public static boolean isDoubleCategory​(ClassNode type)
      It is of a double category, if the provided type is a BigDecimal, a float, double. C(type)=double

    • isFloatingCategory

      public static boolean isFloatingCategory​(ClassNode type)
      It is of a floating category, if the provided type is a a float, double. C(type)=float

    • isNumberCategory

      public static boolean isNumberCategory​(ClassNode type)

    • lowestUpperBound

      public static ClassNode lowestUpperBound​(List<ClassNode> nodes)
      Given a list of class nodes, returns the first common supertype. For example, Double and Float would return Number, while Set and String would return Object.
      Parameters:
      nodes - the list of nodes for which to find the first common super type.
      Returns:
      first common supertype

    • lowestUpperBound

      public static ClassNode lowestUpperBound​(ClassNode a, ClassNode b)
      Given two class nodes, returns the first common supertype, or the class itself if there are equal. For example, Double and Float would return Number, while Set and String would return Object. This method is not guaranteed to return a class node which corresponds to a real type. For example, if two types have more than one interface in common and are not in the same hierarchy branch, then the returned type will be a virtual type implementing all those interfaces. Calls to this method are supposed to be made with resolved generics. This means that you can have wildcards, but no placeholder.
      Parameters:
      a - first class node
      b - second class node
      Returns:
      first common supertype

    • implementsInterfaceOrSubclassOf

      public static boolean implementsInterfaceOrSubclassOf​(ClassNode source, ClassNode targetType)
      Determines if the source class implements an interface or subclasses the target type. This method takes the lowest upper bound class node type into account, allowing to remove unnecessary casts.
      Parameters:
      source - the type of interest
      targetType - the target type of interest