Access.all for maps

[Johanna Larsson]

I recently found the need for, and implemented, a version of `Access.all/0` that handles maps. The implementation is very similar to the one for lists, so I'm curious if anyone has any insight into why there's no built-in version for maps. If there's something I'm missing that makes it difficult to implement in a nice way that doesn't have weird corner cases, or if there just hasn't been a perceived need for it.

I tried searching this mailing list and the Github repo but I wasn't able to spot anything related to "Access.all".

[José Valim]

I don't think there is anything that makes it hard to implement, except the conceptual issues that happen from such feature. If there are duplicate keys, which one wins? The other concern is that the Access API is also mostly specific to one data type, so we would need a new  function for maps and I am not sure if it is worth it. It may be best to roll your own. :)

José Valim

www.plataformatec.com.br

Skype: jv.ptec

Founder and Director of R&D

On Tue, Jun 25, 2019 at 6:26 PM Johanna Larsson <johanna.a...@gmail.com> wrote:

I recently found the need for, and implemented, a version of `Access.all/0` that handles maps. The implementation is very similar to the one for lists, so I'm curious if anyone has any insight into why there's no built-in version for maps. If there's something I'm missing that makes it difficult to implement in a nice way that doesn't have weird corner cases, or if there just hasn't been a perceived need for it.

I tried searching this mailing list and the Github repo but I wasn't able to spot anything related to "Access.all".

--
You received this message because you are subscribed to the Google Groups "elixir-lang-core" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elixir-lang-co...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elixir-lang-core/88ae7e74-5bbb-4b3e-9eaf-cc1344a114e5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Johanna Larsson]

Fair enough! I'm curious what you mean by duplicate keys, that wouldn't happen for a map, would it?

To give more context, I use my version the same as `Access.all/0`:

update_in(

data,

["hosts", &all_map/3, "events", &all_map/3],

updater

)

and the code is pretty much the same, just turn the map into a list before using the same code as `Access.all/0` and then back to map.

On Tuesday, June 25, 2019 at 6:30:11 PM UTC+2, José Valim wrote:

I don't think there is anything that makes it hard to implement, except the conceptual issues that happen from such feature. If there are duplicate keys, which one wins? The other concern is that the Access API is also mostly specific to one data type, so we would need a new  function for maps and I am not sure if it is worth it. It may be best to roll your own. :)

José Valim

www.plataformatec.com.br

Skype: jv.ptec

Founder and Director of R&D

On Tue, Jun 25, 2019 at 6:26 PM Johanna Larsson <johanna....@gmail.com> wrote:

I recently found the need for, and implemented, a version of `Access.all/0` that handles maps. The implementation is very similar to the one for lists, so I'm curious if anyone has any insight into why there's no built-in version for maps. If there's something I'm missing that makes it difficult to implement in a nice way that doesn't have weird corner cases, or if there just hasn't been a perceived need for it.

I tried searching this mailing list and the Github repo but I wasn't able to spot anything related to "Access.all".

--
You received this message because you are subscribed to the Google Groups "elixir-lang-core" group.

To unsubscribe from this group and stop receiving emails from it, send an email to elixir-l...@googlegroups.com.

[José Valim]

Re: duplicates. You could do something like "update_in [:foo, &all_map/3], ..." and that changes the keys in a way you have duplicate keys, which would be lot.

The code snippet look neat though. I guess the more general version of it would be a traverse/3 function powered by a Traverse protocol that knows to enumerate and put things back together. Then it would be able to work with binaries, maps, lists, etc.

José Valim

www.plataformatec.com.br

Skype: jv.ptec

Founder and Director of R&D

To unsubscribe from this group and stop receiving emails from it, send an email to elixir-lang-co...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elixir-lang-core/3598b84f-26a9-4430-b4c3-f449b4174f63%40googlegroups.com.

[Johanna Larsson]

Ah, my version only updates the value, it doesn't expose the key in any way, so keys would always be preserved. Basically, the way that the list version doesn't modify order, just values.

Traverse does sound interesting though! Do you know if any work has been done in that direction?

On Tuesday, June 25, 2019 at 8:53:43 PM UTC+2, José Valim wrote:

Re: duplicates. You could do something like "update_in [:foo, &all_map/3], ..." and that changes the keys in a way you have duplicate keys, which would be lot.

The code snippet look neat though. I guess the more general version of it would be a traverse/3 function powered by a Traverse protocol that knows to enumerate and put things back together. Then it would be able to work with binaries, maps, lists, etc.

José Valim

www.plataformatec.com.br

Skype: jv.ptec

Founder and Director of R&D

[José Valim]

Witchcraft or one of its dependencies does implement Traversal. There was a previous discussion on this list about Empty as well, which could be used to implement traversable with Collectable. But I only mentioned traverse for curiosity. I think all of them would be outside of scope for core. :)

To unsubscribe from this group and stop receiving emails from it, send an email to elixir-lang-co...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elixir-lang-core/28d028b7-374f-487a-b141-6a2c39f4dff1%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

[Johanna Larsson]

Alright, thanks! 

To view this discussion on the web visit https://groups.google.com/d/msgid/elixir-lang-core/CAGnRm4JFmQDEGTFEZor-MHNM8Qm6ABR4zL%3DvsfS6G3GmNdj%2BZg%40mail.gmail.com.