9.2 Formatting ASAM OpenSCENARIO code
9.2.1 Coding style guide
This coding style guide is intended to help you write source code in the ASAM OpenSCENARIO domain-specific language in a uniform way.
9.2.1.1 Introduction
This guide focuses on uniform formatting and naming for source code. It does not contain best practices nor is it a reference for the language.
You should follow the recommendations in this document if you are a scenario developer writing scenarios for real use.
By following the recommendations in this guide the code becomes easy to understand for you and other users.
9.2.2 Formatting
9.2.2.1 Indentation
-
Use four space characters per indentation level.
-
Do not use tab characters.
-
Use the following text editor settings:
-
Tab size: 4
-
Insert spaces
-
9.2.2.3 Whitespace
Whitespace characters provide better readability when used in the correct places.
-
Use only space characters as whitespace.
-
Replace tabs with four spaces as described in Section 9.2.2.1, “Indentation”.
The correct use of space characters within ASAM OpenSCENARIO source code is as follows:
-
One space after a comma (in argument lists and other lists)
-
One space after a colon (for example for inheritance)
-
One space before and after a keyword
-
One space before and after an operator
-
No spaces before and after brackets and braces (for example, in function calls or for indexing)
-
No space between value and unit
The following code snippet shows some examples of the correct use of space characters.
func(arg[1], arg2)
if x == 4: print x, y
foo == [x, y, z]
func(1)
abc[key] = lst[index]
i = i + 1
actor bus: car(category: bus):
keep(width == 1.8m)
keep(length == 4.5m)
swerve_story: serial:
side_vehicle.drive() with:
path(s_side_vehicle)
keep_speed()
with:
until (top.time > 5sec)
9.2.2.4 Comments
-
Single lines of comments and inline comments can be added using the 'pound' symbol (#).
# This is a single line comment.
i = i + 42 # This is an inline comment
-
Blocks of comments can be simulated by using several single lines of comments.
# This is a block of comments example. # The following call to foo is commented out # Reason: see issue xyz # More explanation here... # # foo()
9.2.3 Naming
Naming conventions are a widely discussed topic with great influence on readability influenced by fashion changes, habit, and personal taste.
Here are the recommendations valid for naming conventions in ASAM OpenSCENARIO source code.
9.2.3.1. Use snake_case only
-
Use snake_case (aka lowercase_with_underscore) for all source code elements that are not keywords.
# 1: Define an actor
actor car_group:
average_distance: length
number_of_cars: uint
# 2: Define a road element struct
struct geometric_road: road_element:
min_radius: length
max_radius: length
side: av_side
# 3: Define a scenario
scenario dut.traverse_junction_at_yield:
s: road_with_sign with(sign_type: yield)
do dut.car.traverse_junction() with: ...
# 4: Define a containing scenario
scenario dut.mix_three_dangers:
weather_kind: weather_kind
keep(weather_kind != clear)
do mix:
cut_in_and_slow()
traverse_junction_at_yield()
weather(kind: weather_kind)
9.2.4 Example
Here is a more complex example showing all the rules.
scenario slower_large_vehicle_in_adjacent_lane
ego_vehicle: vehicle with:
keep(p_vehicle_model == APTIV_ego_vehicle)
v1: vehicle with:
keep(vehicle_category == Bus)
keep(p_vehicle_model == BlueBird_Vision_2014)
keep(p_vehicle_model.Color == Blue)
ego_model: ego_vehicle.p_vehicle_model
ego_route: ego_vehicle.route
v1_model: v1.p_vehicle_model
v1_route: v1.route
ego_start_speed: speed
ego_start_distance: length
v1_start_speed: speed
v1_start_distance: length
simulation_platform_choice: string
simulation_time_threshold_reached: time
ego_ttc_threshold: time
ego_distance_to_bias: length
lane_choice: string
ego_lane: lane
map: file_path
keep(map == "/maps/example.xodr")
simple_3_lane_road_01: lane_section
lane1^: lane
lane2^: lane
lane3^: lane
simple_3_lane_road_01.lanes = [lane1^, lane2^, lane3^]
keep(simple_2_lane_road_01 == map.road(id:"1").*lane_section(s:"0"))
lane3^ = rightmost_lane(map.*right)
keep(lane2^ -[:LEFT_OF]-> lane3^)
keep(lane1^ -[:LEFT_OF]-> lane2^)
# Logic parameter syntax
keep(ego_lane in: { if: lane_choice == "Left" THEN: SET ego_lane = "lane1^",
if: lane_choice == "Right" THEN: SET ego_lane = "lane3^"})
keep(ego_model == APTIV_ego_vehicle)
keep(ego_route == ego_drive_left)
keep(v1_model == BlueBird_Vision_2014)
keep(v1_route == v1_drive)
keep(ego_start_speed == 80.46719999999999 [kph])
keep(ego_start_distance == 5 [m])
keep(v1_start_speed == 48.28032 [kph])
keep(v1_start_distance == 55.0 [m])
keep(simulation_platform_choice == "CarMaker")
keep(simulation_time_threshold_reached == 120 [s])
keep(ego_ttc_threshold == 1 [s])
keep(ego_distance_to_bias == 1 [m])
keep(lane_choice == "Left")
# Relationship syntax
keep(v1_start_speed -[:SLOWER_THAN]-> ego_start_speed)
def distance_between_ego_vehicle_and_v1(vehicle, vehicle) -> length is external "kpi.py"
def get_ttc_with_vehicle_to_vehicle(vehicle, vehicle, string) -> time is external "kpi_TTC.py"
!ego_speed: speed = ego_vehicle.speed
!ego_end_of_road: Position = ego_vehicle.end_of_road
!ego_off_road: Position = ego_vehicle.off_road
!simulation_time: time = simulation.time
!ego_vehicle_lane: lane = ego_vehicle.lane
!ego_collision_monitor: integer = sensor.collision.v.fr1.count
!ego_ttc_with_v1: time = sample(ego_vehicle.get_ttc_with_vehicle_to_vehicle(ego_vehicle:vehicle, v1:vehicle, [simulation_platform_choice]))
!ego_distance_to_v1: length = sample(distance_between_egoVehicle_and_v1(v1:vehicle, ego_vehicle:vehicle, [simulation_platform_choice]))
event observation_complete is(one_of(@end_conditions))
event ego_ttc_threshold_reached1 is(!ego_ttc_with_v1 < ego_ttc_threshold)
event ego_close_to_v1 is(!ego_distance_to_v1 < ego_distance_to_bias)
event end_conditions[1] is(!ego_end_of_road > 0)
event end_conditions[2] is(!ego_off_road > 0)
event end_conditions[3] is(!simulation_time > [simulation_time_threshold_reached])
event end_conditions[4] is(!ego_collision_monitor > 0)
event end_conditions[5] is((ego_vehicle.s_road - v1.s_road) > 40)
do parallel():
ego_init_drive_01: ego_vehicle.drive(ego_route) with:
position(at_distance: ego_start_distance, beyond_start_of: ego_route)
speed(at_speed: ego_start_speed)
at(at:start)
v1_init_drive_01: v1.drive(v1_route) with:
position(at_distance: v1_start_distance, ahead_of: ego_vehicle)
speed(at_speed: v1_start_speed)
at(at:start)
until(@observation_complete)
9.2.5 Related topics
If you cannot find a recommendation for your source code formatting or naming problem in this chapter, follow the Style Guide for Python Code (PEP 8).