juancarlospaco
commented
3 months ago !nim c
import std/tables
import std/sequtils
type
Component* = object of RootObj
Subtree[T] = OrderedTableRef[ typedesc[T], seq[T] ]
let universe = newOrderedTable[ string, Subtree[ Component ]]()
proc drop*[T: Component]( uuid: string, components: seq[T] ) =
universe.withValue( uuid, branch ):
branch.withValue( T, leaf ):
for component in components:
leaf.del( component )
proc put*[T: Component]( uuid: string, components: seq[T] ) =
let branch = universe.mgetOrPut( uuid, Subtree[T]() )
if not branch.hasKey( T ):
branch[T] = components
else:
branch.withValue( T, leaf ):
for component in components:
if component notin leaf:
leaf.add( component )
type
Dummy = object of Component
value: int
let d0 = Dummy( value: 0 )
echo $d0
put("zero", @[ d0 ])
github-actions[bot]
metagn
Description
I'm using the playground to try to test out the creation of a simple entity-component relational model in Nim, using generics.
The following code example shows the situation and the test scenario of assigning a single component to an entity.
Nim Version
Using "latest" (via NimPlayground)
Current Output
Expected Output
Possible Solution
No response
Additional Information
As best I can tell from the debugger's output, it's correctly recognizing that I wish to create a
SubtreeofDummy(position 3), but fails to assert that Dummy is an object of Component, despite passing the constraint of[T: Component]to reach line 17.Since the currently available documentation on the nim-lang.org site does not provide guidance on inheritance in generics (or how to explicitly declare this via a type constraint), I would assume that this should resolve normally.
Searching the bug list, this might be related to #88.