Support custom implicit Function Arguments like PosInfos via abstract @:fromNothing macro #6616
Conversation
…ype (unify monos)
…efault value and fromNothing args)
| let infos = mk_infos ctx p [] in | ||
| ordered_args @ [type_expr ctx infos (WithType t)] | ||
| else if ctx.com.config.pf_pad_nulls then | ||
| (ordered_args @ [(mk (TConst TNull) t_dynamic p)]) | ||
| else | ||
| ordered_args | ||
| ordered_args @ [(mk (TConst TNull) t_dynamic p)] |
There was a problem hiding this comment.
@ncannasse does this change, actually ignoring pf_pad_null for bind, does have any effects on real code? The unit tests still pass.
There was a problem hiding this comment.
well it's not ignoring pf_pad_null but is always pushing nulls whereas you're padding or not. I'm not sure I like it, but I guess we can remove extra nulls at a later stage maybe?
There was a problem hiding this comment.
yes, true. I actually wonder if and what implications this actually have.
There was a problem hiding this comment.
or to be more precise, which negative implications this have.
There was a problem hiding this comment.
and just to let you know, there are already cases where this leads to bugs: http://try-haxe.mrcdk.com/#12d04.
|
You will have to rebase this one. |
|
It's up to date again |
|
@ncannasse @Simn @nadako any objections or remarks? |
|
I'm yet to check this out and see what can I do with it :) |
|
I think fromNothing is badly named. I suggest using |
I agree that it's badly named. But |
Yes, it does atm, but its not very useful in my opinion because you don't know the expected type. |
|
What about fields of object literals? Seems like a relatively good use case (will give you prop drilling for example). Would there be anything wrong with also using this for initializing variables? |
|
@back2dos would love to see an example, because i fail to see how that could be done without having an abstract with a concrete underlying type. |
|
Let's revisit this for 4.1 |
|
This branch would need an update. @RealyUniqueName Could you look into this? |
|
Sure |
|
This PR is unfortunately unmergeable at this point. I'll link it to #4583 mostly because there are some good tests here, so if we add this feature for Haxe 5 we can add those in some way. |
No description provided.