Route progress classes should be codable

[1ec5]

RouteProgress, RouteLegProgress, and RouteStepProgress should conform to NSSecureCoding or Codable. MapboxDirections.swift classes such as Route are already NSSecureCoding-conformant.

/cc @mapbox/navigation-ios

[1ec5]

Once we move the progress classes off NSObject (#2294), it’ll be straightforward to implement Codable conformance in these classes. This will make it much easier for developers to implement state restoration.

[Udumft]

RouteLegProgress already conforms to Encodable here:

mapbox-navigation-ios/MapboxCoreNavigation/EventDetails.swift

Lines 298 to 329 in 134af22

	extension RouteLegProgress: Encodable {
	private enum CodingKeys: String, CodingKey {
	case upcomingInstruction
	case upcomingType
	case upcomingModifier
	case upcomingName
	case previousInstruction
	case previousType
	case previousModifier
	case previousName
	case distance
	case duration
	case distanceRemaining
	case durationRemaining
	}
	public func encode(to encoder: Encoder) throws {
	var container = encoder.container(keyedBy: CodingKeys.self)
	try container.encodeIfPresent(upcomingStep?.instructions, forKey: .upcomingInstruction)
	try container.encodeIfPresent(upcomingStep?.maneuverType, forKey: .upcomingType)
	try container.encodeIfPresent(upcomingStep?.maneuverDirection, forKey: .upcomingModifier)
	try container.encodeIfPresent(upcomingStep?.names?.joined(separator: ";"), forKey: .upcomingName)
	try container.encodeIfPresent(currentStep.instructions, forKey: .previousInstruction)
	try container.encode(currentStep.maneuverType, forKey: .previousType)
	try container.encode(currentStep.maneuverDirection, forKey: .previousModifier)
	try container.encode(currentStep.names?.joined(separator: ";"), forKey: .previousName)
	try container.encode(Int(currentStep.distance), forKey: .distance)
	try container.encode(Int(currentStep.expectedTravelTime), forKey: .duration)
	try container.encode(Int(currentStepProgress.distanceRemaining), forKey: .distanceRemaining)
	try container.encode(Int(currentStepProgress.durationRemaining), forKey: .durationRemaining)
	}
	}

But this implementation does not seem to be exactly what is expected, because it does not directly encodes it's variables and instead encodes sub-sub properties. I thought that it is intended to be used for passing as context info for some Events, but I could not find any real usage.
What is this implementation for?

[JThramer]

Blocked by #2352

[JThramer]

Posterity: Regarding this comment, as long as route-step is encodable, RouteLegProgress should just encode RouteStep directly. Needs investigation.

[zugaldia]

Closing per sprint planning convo.

[1ec5]

Sorry, I got some wires crossed in sprint planning. This change is still relevant but doesn’t need to happen for v1.0.0.

But this implementation does not seem to be exactly what is expected, because it does not directly encodes it's variables and instead encodes sub-sub properties. I thought that it is intended to be used for passing as context info for some Events, but I could not find any real usage.
What is this implementation for?

Looks like this implementation was introduced in 66cb213 for #1533 but ended up being unused after 982a94c in the same PR. We can safely remove that implementation and replace it with a more standard one that directly encodes its members.