>From what I understood from the previous messages & discussion on GitHub, it
>would be more convenient for users to be able to have a List instead of an
>Iterable, or instead of having to call the #toList() or convert to a List in
>some other way.
I commented in the pull request, that I don't think there would be a
performance penalty in doing so (at least I don't think so, as the values are
not streamed, but rather kept in the private values array).
However, I think we are delivering an Iterable that's fully capable to be used
as an Iterable now. Whereas the proposal would make it a read-only list, as
that returned from unmodifiableList, i.e. throwing exceptions for add/clear/etc
operations.
In my opinion, I prefer to keep it as an Iterable, leave the toList method, as
I think current users could be affected by accidentally trying to reuse
CSVRecord while reading from one input and writing to an output stream.
So I'm -0 for it.
Bruno
From: sebb <[email protected]>
To: Commons Developers List <[email protected]>
Sent: Tuesday, 18 December 2018 12:24 AM
Subject: Re: [CSV] Records as Lists
What is the use-case for using lists?
On Thu, 13 Dec 2018 at 18:34, Gary Gregory <[email protected]> wrote:
>
> Hi All,
>
> I am looking for opinions on turning a CSV record into a list, as opposed
> to the minimal current implementation. There would be side-effects like a
> record becoming writable instead of read-only as the current implementation.
>
> Memory footprint would also be a concern.
>
> Please see https://github.com/apache/commons-csv/pull/35
>
> Gary
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
