Dietrich Rothe authored 2 years ago and Dietrich Rothe committed 2 years ago

and migrate it to use CDD syntax

  DeviceModule and ControlSystemModule now do not take the optional prefix argument in their constructors any more. Use the [] operator instead to get the submodule.

- Add ExceptionHandlingDecorator and decorate all device variables with it.

Use one file per class. Also, use boost::core::demangle instead of
checking for abi::__cxa_demangle by ourselves.

moved the xml generator main function into a new header file, to avoid defining main() multiple times by mistake.

A DeviceModule now returns the same VariableNetworkNode when asked for the same register multiple times. This leads to a new restriction that registers can now only be either read or written. This restriction can be mitigated by using the logical name mapper to create an alias of the register. This has been implemented for the tests and the demo application.

- allow setting the "eliminateHierarchy" flag in the constructor of a module

This is an interface change which was done in an incompatible way!

Use mtca4u::setDMapFilePath() instead of the BackendFactory's member function. Since all applications will need it, the corresponding header file is now included in ApplicationCore.h

Removed the macros for defining accessors and introduced convenience classes instead. This adds slightly more syntactic sugar, but also allows more flexibility and seems to be a more coherent interface.

This commit incompatibly breaks the interface with the previous versions. Since ApplicationCore is not yet used in any productive environment, no compatibility is provided.

updated the example: use "using" syntax to inherit constructors and re-enabled the previously not working read-back of the device register "Variable".

Added names and owners for modules, to build a full hierarchy chain. The syntax for creating modules is now not so nice, a better way has to be found (maybe by adding a macro if better solution can be found).

Read all application inputs once before starting the module threads, so the initial values are correctly present.

For this purpose, additional changes were necessary:
- Added support for constant values as feeders and consumers (just ignore the writes)
- Connect all unconnected application accessors to constants after defining the connections and before creating the implementations
- Renamed the user-overridden function "initialise()" into "defineConnections()", since more actions are now necessary during initialise().
- The connections are now made (i.e. the implementations are created) during the initialisation phase

Added description field to the variables. It is currently ignored and will be added to the XML file in a next step.

reading back the variable from the device currently does not work, so instead directly pass it to the simulator module.

- extended the trigger tests
- removed some debug output

Introduced the DeviceModule class which provides an interface to obtain VariableNetworkNodes for device registers

- performed some code cleaning and refactoring
- store all nodes in a VariableNetwork in a single list, to allow deciding later if a node is feeding or consuming

This commit is an intermediate commit and has known bugs. Almost all tests are failing and the demoApp is not running. Debugging will be done in a next step.

introduced an operator>> and operator<< syntax to connect network nodes. to realise this, the VariableNetworkNode class has been changed into a pimpl pattern.

Allow to make connections directly based on VariableNetworkNodes. This is a first step, more changes have to be made in this direction.

Split the SCALAR_ACCESSOR macro into SCALAR_OUTPUT and SCALAR_INPUT to prevent output accessors to be created in poll mode (which makes no sense). Since some tests were using this, now instead a DeviceAccessor in poll mode is used for these tests.

- fixed a bug in the FanOut which did not wait for the trigger (since ProcessScalar::receive() does not wait)
- fixed a bug in Application::feedDeviceRegisterToControlSystem() not handling the last arugment (trigger) properly
- improved the output produced with VariableNetwork::dump()