It really sounds like csv should just be a format that's plugged into
another library.
On Nov 29, 2011 9:29 AM, "Emmanuel Bourg" <ebo...@apache.org> wrote:
> Le 29/11/2011 14:43, Matt Benson a écrit :
>
> Well, assuming "header-free" CSV output you could do any odd thing like:
>>
>> foo;bar;(2);element1;element2;
>>
>> giving an open-ended format. Not saying such would be the greatest
>> idea, but could be usable under the right circumstances.
>> Alternatively, one could embed the collection:
>>
>> foo;bar;"{element1;element2}"
>>
>
> ... and it starts to get fun when you want to escape the special
> characters '"', ';' and '}' in the collection ;)
>
> It's possible to tailor a CSV like format that supports collections, but
> if it doesn't interoperate with other CSV tools it's hardly useful. If
> people want a more capable format without caring about interoperability
> there are legions of alternative serialization APIs available.
>
> That's why I think [csv] should focus on the simple format defined by
> RFC4180, exotic features are for more ambitious projects (at least for now).
>
> Emmanuel Bourg
>
> ------------------------------**------------------------------**---------
> To unsubscribe, e-mail:
> dev-unsubscribe@commons.**apache.org<dev-unsubscr...@commons.apache.org>
> For additional commands, e-mail: dev-h...@commons.apache.org
>
>
