The DPSMF searches a project for yaml files with a certain naming convention and generates boilerplate APIs for each of the files found. The boilerplate text, as the project name implies, is for publishers and subscribers.

• Publishers: Code that creates data and wants to share it
• Subscribers: Code that wants to get data from a publisher

Along with the boilerplate code, a document outlining the API created will also be generated at the location specified by the out_dir option. There are two main targets for this project:

• Simulation
• Actual implementation framework (TBD)

To initialize the program, the base_path must be specified from where you wish the program to begin recursively searching for appropriately named YAML files. The naming convention for the YAML file goes as follows: [type]_[name].yml. As an example, if one wishes to create a publisher called position, the required file name would be pub_position.yml.

Full documentation can be found here

TODO: General formatting

The following section contain explanations for the parameters of each of the allowed types.

• Variable/Parameter API
  • Int
  • Float
  • String
  • Boolean
  • Arrays

Publishers, subscribers, and system YAML files define the parameters required to generated instances for each of their respective types. At a high level, publishers provide data to the system, subscribers listen to the published data, and system files are general purpose glue code to interface with hardware drivers and simulation.

The figure below describes how publishers interact with DPSMF. When a publisher is created, the parameters specified by the publisher are used too

1. generate subscriber code in the directory the YAML file was found and
2. create an internal representation of the data structure in DPSMF.

The file generated provide the necessary code to be able to publish whatever information its provided to the DPSMF client. Once the data has been published to the client (whether it be a local or remote client), it is the job of the DPSMF client too

1. store the data in a database and
2. transmit the data utilizing the specified mode of transmission (TCP, UDP, RF, etc.)

The database can be transient in the sense that it only keeps the latest information (i.e., used to reduce storage footprint), or it can be persistent to store specific or all data provided for logging and playback purposes.

• Type API
  • Publisher API
  • Subscriber API
  • System API

The simulation aspect will generate the publishers, subscribers, and mock frameworks. By default DPSMF will generate just the publishers and subscribers, meaning that the mock interfaces will be generated only if the enable_mock option is set to true.

• Python
• Rust
• C/C++