dsv.format could take an optional columns array? about d3-dsv

[mbostock]

This is why there’s a dsv.formatRows method. I’ve added some examples to the API reference. Hope this helps!

[mariocasciaro]

Thanks for the answer, it's all clear now!

[mbostock]

Even though you can do this with dsv.formatRows, I think it would be easy to support this, so I’m reopening this issue. As you suggested, the dsv.format method could take an optional array of column names. Then, instead of this:

var string = csv.formatRows([[
    "year",
    "make",
    "model",
    "length"
  ]].concat(data.map(function(d, i) {
  return [
    d.year,
    d.make,
    d.model,
    d.length
  ];
})));

You could say this:

var string = csv.format(data, ["year", "make", "model", "length"]);

Which is obviously much nicer. 😁

Another possibility is that we could check if the passed-in rows array exposes a columns array, now that dsv.parse supports this feature. But I think this is probably too magical, since you might forget about the existence of the columns property after transforming the array. So, better to make it an explicit argument.

[mbostock]

Fixed in 0.1.11.

[mariocasciaro]

Thanks for the feature! It's definitely a much neater approach