Remove one of two functions with same implementation, generally make consistent names

[jvoigtlaender]

deleteIf and removeWhen are the same function. I think the library should contain only one of them.

[jvoigtlaender]

Also, there's "removeAt", which might as well be called "deleteAt". The library could use a pass of making names consistent.

[jvoigtlaender]

The last comment, because there are also a bunch of "drop..." functions. What is the difference between deleting, removing, and dropping?

[shmookey]

Edit: now that this issue is closed, I've cleaned up the following list to make it easier to refer back to.

If the goal is to go for consistent naming, then to me:

• filters operate over all elements of a list
• drops operate on a list prefix, conversely dropRights operate on a suffix
• removes operate on the first matching element[s] of a list
• deletes are not yet defined, this term is available for some other class of behaviour
• before suggesting that a function which is not a filter, drop or remove should be named 'delete', consider using a name that describes what you get back rather than what the function does, e.g. instead of 'dropDuplicates', use 'unique'

My deleteFirst (https://github.com/elm-community/list-extra/pull/6) operates on the first matching element, but that's just copying a Haskell convention. To my ear "delete" connotes erasing data from persistent storage, rather than merely discarding it from a data structure in memory. I'd prefer to clean up the other names and rename it to remove.

I would suggest:

• Start considering "remove" to suggest operations on the first matching element of a list.
• Keep removeAt because it fits this constraint.
• Rename removeWhen to filterNot (or drop it entirely because it's trivial).
• Delete deleteIf because it's the same as filterNot.
• Accept my PR, but rename deleteFirst to remove, which is consistent with Set.remove

[jvoigtlaender]

What about the function currently called dropDuplicates?

And I assume there would be renaming of current dropWhileEnd to dropWhileRight, and analogously for takeWhileEnd?

[shmookey]

I think that makes sense. Re: dropDuplicates, I'm inclined to think of it in terms of what you get back - how about uniques?

Happy to put these changes in my PR.

[mgold]

In my nonempty list library, I offer uniq and dedup. The former removes all duplicates, while the latter removes adjacent duplicates. I'm not thrilled with the abbreviations, but maybe unique, as a verb, rather than the plural noun?

[shmookey]

Yeah, sounds good. Updated #6.

[jvoigtlaender]

Some context about dropDuplicates/unique: https://github.com/elm-community/elm-list-extra/issues/31.

Also, a change here affects https://github.com/elm-community/list-extra/pull/5, so bring this name change up there as well?