Use … to incidate there are hidden rows

[billylanchantin]

In discord we chatted about using … to indicate that there are hidden rows. This PR does that, but I ended up doing a few other things while I was at it.

This PR does:

1. Adds a row of … to indicate there are hidden rows. Includes a new option limit_dots: :bottom | :split to specify how to do this.
2. Drops the row separators when they aren't needed (see below).
3. Allows you to pass through valid options to TableRex.render!/2. This gives you a little more flexibility in case you don't like the defaults.

Example

Explorer.Datasets.iris() |> Explorer.DataFrame.print()

Before:

+-----------------------------------------------------------------------+
|              Explorer DataFrame: [rows: 150, columns: 5]              |
+--------------+-------------+--------------+-------------+-------------+
| sepal_length | sepal_width | petal_length | petal_width |   species   |
|    <f64>     |    <f64>    |    <f64>     |    <f64>    |  <string>   |
+==============+=============+==============+=============+=============+
| 5.1          | 3.5         | 1.4          | 0.2         | Iris-setosa |
+--------------+-------------+--------------+-------------+-------------+
| 4.9          | 3.0         | 1.4          | 0.2         | Iris-setosa |
+--------------+-------------+--------------+-------------+-------------+
| 4.7          | 3.2         | 1.3          | 0.2         | Iris-setosa |
+--------------+-------------+--------------+-------------+-------------+
| 4.6          | 3.1         | 1.5          | 0.2         | Iris-setosa |
+--------------+-------------+--------------+-------------+-------------+
| 5.0          | 3.6         | 1.4          | 0.2         | Iris-setosa |
+--------------+-------------+--------------+-------------+-------------+

After:

iex> Explorer.Datasets.iris() |> Explorer.DataFrame.print()
+--------------------------------------------------------------------------+
|               Explorer DataFrame: [rows: 150, columns: 5]                |
+--------------+-------------+--------------+-------------+----------------+
| sepal_length | sepal_width | petal_length | petal_width |    species     |
|    <f64>     |    <f64>    |    <f64>     |    <f64>    |    <string>    |
+==============+=============+==============+=============+================+
| 5.1          | 3.5         | 1.4          | 0.2         | Iris-setosa    |
| 4.9          | 3.0         | 1.4          | 0.2         | Iris-setosa    |
| 4.7          | 3.2         | 1.3          | 0.2         | Iris-setosa    |
| …            | …           | …            | …           | …              |
| 6.2          | 3.4         | 5.4          | 2.3         | Iris-virginica |
| 5.9          | 3.0         | 5.1          | 1.8         | Iris-virginica |
+--------------+-------------+--------------+-------------+----------------+

Row separators

The row separators have bothered me for a while. They effectively double the vertical size of the printed dataframe. But they're only necessary when there are multi-line cells.

With this PR, I choose to only include row separators when there's a "composite" dtype (list or struct). IMO this should cover most cases where the extra row separator is beneficial in a predictable way.

Another way to do it is to detect the presence of multi-line cells. There are tradeoffs to both approaches.

I'm also happy to scale this PR back to just the … part.