Computes the information for the given project.
Computes the information for the given project.
Classes that inherit from this trait are not expected to
make this method public. This method is only expected to be called
by an instance of a Project.
Returns the information which other project information need to be available before this analysis can be performed.
Returns the information which other project information need to be available before this analysis can be performed.
If the analysis has no special requirements Nil can be returned.
Classes/Objects that implement this trait should not make the method public
to avoid that this method is called accidentally by regular user code.
All requirements must be listed; failing to specify a requirement can end up in a deadlock.
The unique id of this key.
The unique id of this key. The key is used to enable efficient access and is automatically assigned by OPAL and will not change after that.
ProjecInformationKeyobjects are used to get/associate some (immutable) information with a project that should be computed on demand. For example, imagine that you write an analysis that requires – as a foundation – the project's call graph. In this case, to get the call graph it is sufficient to pass the respective key to the Project object. If the call graph was already computed that one will be returned, otherwise the computation will be performed and the result will be cached for future usage before it is returned.Using Project Information
If access to some project information is required it is sufficient to use the (singleton) instance of the respective
ProjectInformationKeyto get the respective project information.For example, let's assume that an index of all fields and methods is needed. In this case the code to get the index would be:
Providing Project Information/Implementing
ProjectInformationKeyMaking project wide information available on demand is done as follows.
ProjectInformationKeyclass that inherits from this trait and which calls the base analysis. It is recommended that the factory method (compute) is side-effect free.Threading
Project takes care of threading related issues. The methods requirements and compute will never be called concurrently w.r.t. the same
projectobject. However, concurrent calls may happen w.r.t. two different project objects.Caching
Project takes care of the caching of the result of the computation of the information.