Create abstractions for creating common shacl shapes/constraints

[TShapinsky]

Example:

ahu_feeds_vav = ConnectionConstraint(BRICK["AHU"], BRICK["feeds"], BRICK["VAV"])
shape_collection.add_graph(CountConstraint(ahu_feeds_vav, 5))
shape_collection.add_graph(CountConstraint(BRICK["VAV"], 5))

[TShapinsky]

@MatthewSteen @gtfierro @haneslinger Thoughts?

# Using generic rdf shape builder
vav_shape = ShapeBuilder('simple_vav')
  .of_class(BRICK['VAV'])
  .has_relationship(BRICK['isFedBy'],target_class=BRICK['AHU'], count=1)
  .has_relationship(BRICK['feeds'],target_class=BRICK['Zone'], min=1)
  .has_relationship(BRICK['hasPoint'],target_class=BRICK['Zone_Temperature_Sensor'], count=1)
  .make()
  
ahu_shape = ShapeBuilder('simple_ahu')
  .of_class(BRICK['AHU'])
  .count(5)
  .has_relationship(BRICK['feeds'],target_shape=vav_shape, count=5)
  .has_relationship(BRICK['hasPoint'], target_class=BRICK['Outdoor_Air_Temperature_Sensor'], count=1)
  .has_relationship(BRICK['hasPoint'], target_class=BRICK['Return_Air_Temperature_Sensor'], count=1)
  .make()
  
manifest = vav_shape + ahu_shape

# ShapeBuilder extended for Brick specifically
vav_shape = BrickShapeBuilder('simple_vav')
  .of_class(BRICK['VAV'])
  .is_fed_by(BRICK['AHU'])
  .feeds(BRICK['Zone'], min=1)
  .has_points([BRICK['Zone_Temperature_Sensor'], BRICK['Zone_Temperature_Setpoint']], count=1)
  .make()
  
ahu_shape = BrickShapeBuilder('simple_ahu')
  .of_class(BRICK['AHU'])
  .count(5) # Should be 5 in the model
  .feeds(vav_shape, count=5)
  .has_points([BRICK['Outdoor_Air_Temperature_Sensor'],BRICK['Return_Air_Temperature_Sensor']], count=1)
  .make()
         
manifest = vav_shape + ahu_shape

[MatthewSteen]

I like the concept and think it would be a helpful feature. How would the error checking work for all the possibilities of arguments for the different ontologies of 223, Brick, Haystack? Some of this seems reminiscent of https://github.com/gtfierro/oomason.