Re: CSV+ file lines with differing number of columns

On 19 February 2014 06:06, Ivan Herman <ivan@w3.org> wrote:
>
>
> Tim Finin wrote:
>> The current draft of Syntax for Tabular Data on the Web
>> stipulates (sec 3.3) that "Each line of a CSV+ file must contain
>> the same number of comma-separated values."  While this seems
>> reasonable, some existing use cases I'm familiar with allow for
>> CSV files with several types of lines that differ in their number
>> of columns.  Processing the CSV file requires detecting the line
>> type and also the presence of an optional terminal column.
>>
>> Might we explore relaxing the constraint that the CSV file have
>> the same number of columns for each line?
>
> Actually... we have to be careful how we formulate all this. The working group
> is not in position to define or not constraints, because we are not in position
> to define what CSV is. The only thing we can do is to describe what is out
> there, and adapt our output accordingly...
>
> (Just with my administrative hat on...)

We have some wiggle-room. It is quite reasonable for us to say "this
is the subset of 'wild CSV' to which we consider our work applicable."

The charter http://www.w3.org/2013/05/lcsv-charter also says
"Contributing to a new version of RFC 4180" in collab w/ IETF is in
scope. And nothing says that RFC-4180 is ground truth w.r.t. what
counts as CSV. If it has data in it separated by ',', someone
somewhere probably calls it CSV. It's a messy world out there!

Dan

Received on Wednesday, 19 February 2014 17:29:56 UTC