Configuration

The configuration system allows to define named values with defaults that can be changed and queried while the secondary device is running with a telecommand. The system has two components.

Configuration definition

The configurations are defined in the configuration.csv file. It consists of a csv table with four columns and one header row, which contains the name of the columns. Each row represents one configuration item:

Column name	Column description
Name	The name of the configuration item.
Type	The type that is used to represent this configuration item. Can be a unit or shared data type.
Default Value	The default value of this configuration item.
Description	An optional human readable description of this configuration item.

The following example defines two configuration items: A boolean imu enabled, which is enabled by default, and a sensor check interval, which is 5000 milliseconds by default, using a ms unit.

Name,Type,Default Value,Description
imu enabled,bool,true,Wheter the IMU is enabled.
sensor check interval,ms,5000,How much time to wait between checking the sensor.

Using the configuration items specified in the configuration.csv file, the database will automatically add the following constants:

NUM_CONFIGURATIONS: The number of provided configuration items. In the example above, it would be the number 2.
DEFAULT_CONFIGURATION: A struct that contains the default configuration of the declared config items. For the example above, represented as a json object, it would be
```
{
  "imu enabled": true,
  "sensor check interval": 5000
}
```
MAX_CONFIG_VALUE_SIZE: The maximum size in bytes of any configuration value. In the example above, it would be the maximum between 1 (the size of bool) and the size of the ms unit.

Additionally, the following types will be defined:

ConfigurationId: An enum type containing an enum value for each configuration item.
Configuration: A struct type that contains a child for each configuration item. This is the type used in the DEFAULT_CONFIGURATION constant, and it can be used to hold a configuration state.

Configuration telecommands

If a command is defined with one argument of type ConfigurationId and a return value of the special type config?, the database will mark this command as a configuration getter telecommand. An automatic handler for this telecommand will be generated in the C generator, which responds to this telecommand with the value of the configuration indicated by the ConfigurationId argument. A ResponseParser is able to automatically convert the response to the correct data type. The automatically generated C code requires the following function to be defined somewhere and to provide access to the current configuration:

const Configuration* getConfig();

Also, if a command is defined with no return types and two arguments, one of type ConfigurationId and another one of the special type config?, the database will mark this command as a configuration setter telecommand. An automatic handler for this telecommand will be generated in the C generator, which correctly parses the configuration value to its correct data type and updates the configuration with the new value. The automatically generated C code requires two functions to be defined somewhere:

A function to get a mutable configuration object and indicate that a change will be made.

Configuration* beginConfigUpdate();

And a function to commit the changes made to the mutable configuration object retrieved with beginConfigUpdate.

void commitConfigUpdate();