delete function vsremove function?

delete functions returns the deleted element.
remove function does same job as todelete function but doesn't return the deleted element.

Any suggestions on better naming convention? Should I rather calldelete aspop and think of a better name forremove?

This is highly subjective but FWIW, I would preferpop for the functions that return the deleted element, simply because I imagine that if I "pop" something out I still have it with me.

For the function that doesnot return the deleted element I am fine with bothremove ordelete, with a slight preference fordelete.

Did you have any time to see what are the choices of other languages for similar functions?

Python usespop (returns deleted item) andremove (no return)
Java doesn't have any function equivalent ofpop (or atleast I couldn't find one), it hasremove function (no return).

pop function is used whenever deleted item is returned, otherwiseremove anddelete functions both are used by programming languages. Though, I could find only one occurrence wheredelete was used to remove elements.
Also, some languages usespop() (with no arguments) as a way to allow a user to make alist behave like astack or aqueue.

I thinkpop is a reasonable name to a procedure to remove and return the removed element. For only deleting elements,delete,remove ordrop might be suitable. The namesdelete anddestroy could be associated with a deconstructor of the entire list.

Based on my experience with past discussions, we are trying to provide a familiar interface to the user ofstdlib following well-established conventions from other languages (like python, for example).@arjenmarkus suggests more "Fortranic" names (that I like too) but they might be inconsistent with other parts ofstdlib.

Based on the other suggestions, my preference is

• pop: removes and returns the removed element
• drop: removes an element without returning it
• destroy: de-constructs the list and frees the memory

I do not have strong opinions on this, though.

I agree withdestroy (please note that we have a function namedclear instdlib_stringlist which resets the list and NOT destroys it). Checkout this example frompython to understand the difference:

I added a basic range feature topop anddrop which will remove all strings at indexes in the interval [first,last].
Out of bounds indexes in this interval will be ignored. There is no concept ofstride currently in this range feature.

We can discuss the behaviour of non-basic range feature forpop anddrop.

I think atleast a basic range feature should be added toget function as well, it will also improve the waypop_engine function is written.

Thank you.

I personally don't like that out of bounds indexes are ignored: I'd like to be notified (at least with a warning, if not by an error) in these cases. Based on my experience, out of bounds indexes are the consequence of upstream bugs that would be difficult to identify, if ignored. However, I believe that a choice in this sense might have been already made during the initial implementation ofstringlist for other methods.

I think atleast a basic range feature should be added toget function as well, it will also improve the waypop_engine function is written.

Good point, I agree.

I am under the impression that adding theget function (although, as I said, it's a good idea) is out of scope for this PR. Am I right?

I am under the impression that adding theget function (although, as I said, it's a good idea) is out of scope for this PR. Am I right?

Yeah, that's right. Let me rename PR.

This PR is blocked by:#552