named tuple toMap extension method

[road21]

There already are methods toArray, toList for NamedTuple for collecting elements into collections.
I would like to add method toSeqMap where keys are names of the elements.

For example:

(x = 1, y = true).toSeqMap // SeqMap("x" -> 1, "y" -> true)

One obvious usage of this method is debugging code with named tuples (standard toString method doesn't show names):

println((x = 1, y = true)) // prints (1, true)
println((x = 1, y = true).toSeqMap) // prints SeqMap(x -> 1, y -> true)

[He-Pin]

Will there be a toListMap

[road21]

Will there be a toListMap

I think toListMap is enough, fixed to toListMap

[road21]

@aherlihy could you please take a look?

[aherlihy]

@road21 Is this ready for review? Once its rebased I will take a look.

[aherlihy]

Also, at first glance, you should abstract over the implementation and return a Map (not ListMap), and name it toMap, as we do not want to add something to the standard library and be required to keep the implementation forever.

[road21]

@road21 Is this ready for review? Once its rebased I will take a look.

Yes, rebased

Also, at first glance, you should abstract over the implementation and return a Map (not ListMap), and name it toMap, as we do not want to add something to the standard library and be required to keep the implementation forever.

What about toSeqMap: SeqMap[String, Tuple.Union[V]] then?
I would like to keep the order of elements

[road21]

@road21 Is this ready for review? Once its rebased I will take a look.

Please, take a look. I have fixed method to SeqMap

[bracevac]

LGTM. The failing CI test is unrelated. Please also squash the commits. It's a small change set.

[hamzaremmal]

I would like to keep the order of elements

May I ask why?

[road21]

May I ask why?

Just because we have an order and I don't see any downsides of toSeqMap instead of toMap.
If user will need list of pairs (field name, field value) - toSeqMap fits.
I don't have any realistic example yet.
But I can change it back to toMap if you wish.

[bracevac]

It's perfectly reasonable to preserve the order. Unless @hamzaremmal has more specific concerns, I think this is ready. The CI problems have been resolved apparently. A rebase is probably needed.

[hamzaremmal]

It's perfectly reasonable to preserve the order. Unless @hamzaremmal has more specific concerns, I think this is ready. The CI problems have been resolved apparently. A rebase is probably needed.

I agree that preserving the order is reasonable. I have some use cases where this is important and I was just curious what would be @road21's use case.