In general an IsReferenceValue
abstracts over all potential values and this information is
sufficient for subsequent analyses; but in some cases, analyzing the set of underlying values
may increase the overall precision and this set is returned by this function.
In general an IsReferenceValue
abstracts over all potential values and this information is
sufficient for subsequent analyses; but in some cases, analyzing the set of underlying values
may increase the overall precision and this set is returned by this function. In other
words: if baseValues
is nonEmpty, then the properties return by this
value are derived
from the base values, but still maybe more specific. For example,
Object o = _; if(...) o = f() else o = g(); // when we reach this point, we generally don't know if the values returned by f and g // are non-null; hence, o is potentially null. if(o != null) // Now, we know that o is not null, but we still don't know if the values returned // by f OR g were null and we cannot establish that when we don't know to which value // o is actually referring to. u(o);
the set of values this reference value abstracts over. The set is empty if this value is already a base base value.
A reference value which belongs to the base values by some other reference value never has base values itself.
The upper bound of the value's type.
The upper bound of the value's type. The upper bound is empty if this
value is null
(i.e., isNull == Yes
). The upper bound is only guaranteed to contain
exactly one type if the type is precise. (i.e., isPrecise == true
). Otherwise,
the upper type bound may contain one or more types that are not known to be
in an inheritance relation, but which will correctly approximate the runtime
type.
If only a part of a project is analyzed, the class hierarchy may be fragmented and it may happen that two classes that are indeed in an inheritance relation – if we would analyze the complete project – are part of the upper type bound.
The set of base values this values abstracts over.
The set of base values this values abstracts over. This set is never empty and contains this value if this value does not (further) abstract over other reference values.
Primarily defined as a convenience interface.
If Yes
the value is known to always be null
at runtime.
If Yes
the value is known to always be null
at runtime. In this
case the upper bound is (has to be) empty. If the answer is Unknown
then the
analysis was not able to statically determine whether the value is null
or
is not null
. In this case the upper bound is expected to be non-empty.
If the answer is No
then the value is statically known not to be null
. In this
case, the upper bound may precisely identify the runtime type or still just identify
an upper bound.
This default implementation always returns Unknown
; this is a sound
over-approximation.
Unknown
.
This method is expected to be overridden by subtypes.
Returns true
if the type information is precise.
Returns true
if the type information is precise. I.e., the type returned by
upperTypeBound
precisely models the runtime type of the value.
If, isPrecise
returns true, the type of this value can
generally be assumed to represent a class type (not an interface type) or
an array type. However, this domain also supports the case that isPrecise
returns true
even though the associated type identifies an interface type
or an abstract class type. The later case may be interesting in the context
of classes that are generated at run time.
This default implementation always returns false
.
false
This method is expected to be overridden by subtypes.
,isPrecise
is always true
if this value is known to be null
.
True in case of a value with primitive type; undefined if the type is unknown.
True in case of a value with primitive type; undefined if the type is unknown.
True if the value has a reference type; undefined if the type is unknown.
True if the value has a reference type; undefined if the type is unknown.
Returns true
if no type information is available.
Returns true
if no type information is available.
Tests if the type of this value is potentially a subtype of the specified
reference type under the assumption that this value is not null
.
Tests if the type of this value is potentially a subtype of the specified
reference type under the assumption that this value is not null
.
This test takes the precision of the type information into account.
That is, if the currently available type information is not precise and
the given type has a subtype that is always a subtype of the current
upper type bound, then Unknown
is returned. Given that it may be
computationally intensive to determine whether two types have a common subtype
it may be better to just return Unknown
in case that this type and the
given type are not in a direct inheritance relationship.
Basically, this method implements the same semantics as the ClassHierarchy
's
isSubtypeOf
method, but it additionally checks if the type of this value
could be a subtype of the given supertype. I.e., if this value's type
identifies a supertype of the given supertype
and that type is not known
to be precise, the answer is Unknown
.
For example, assume that the type of this reference value is
java.util.Collection
and we know/have to assume that this is only an
upper bound. In this case an answer is No
if and only if it is impossible
that the runtime type is a subtype of the given supertype. This
condition holds, for example, for java.io.File
which is not a subclass
of java.util.Collection
and which does not have any further subclasses (in
the JDK). I.e., the classes java.io.File
and java.util.Collection
are
not in an inheritance relationship. However, if the specified supertype would
be java.util.List
the answer would be unknown.
This default implementation always returns Unknown
.
This method is expected to be overridden by subtypes.
,The function isValueSubtypeOf
is not defined if isNull
returns Yes
;
if isNull
is Unknown
then the result is given under the
assumption that the value is not null
at runtime.
In other words, if this value represents null
this method is not supported.
Characterizes a reference value. Captures the information about the values a domain value may refer to. For example, in the following:
o is a reference value (
IsReferenceValue
) that (may) refers to two "simple" base values:new Object()
and"STRING"
; however, it is a decision of the the underlying domain whether the information about the base values is made available or not. Furthermore, if the base values are actually used, the constraints in effect for the overall abstraction should be considered to get the most precise result.