[C++] Add support for specifying custom Array element delimiter to `arrow::PrettyPrintOptions`

[kevingurney]

Describe the enhancement requested

In order to make the arrow::PrettyPrint functionality for arrow::Array more flexible, it would be useful to be to specify a custom Array element delimiter other than ",".

For example, the MATLAB interface wraps the Arrow C++ libraries and being able to specify a custom Array element delimiter, would make it possible to make the display of MATLAB arrow.array.Array objects more MATLAB-like.

In order to support custom Array element delimiters, we could add a new option to the arrow::PrettyPrintOptions struct named something like array_element_delimiter with type std::string.

Implementing this functionality would enable use cases like the ability to display an arrow::Array that normally displays as [1,2,3] to display as [1 | 2 | 3] by setting array_element_delimiter = " | ".

Component(s)

C++

[rok]

This is related but stalled: #30951.

[kevingurney]

@rok - thank you for sharing this! My apologies - I didn't see this issue before.

I just opened #37981 to add support for specifying a custom Array element delimiter. If I am not mistaken, it looks like #30951 took the approach of trying to change the default delimiter to be ", " instead of just ",". My approach is slightly different in that it adds a new property to arrow::PrettyPrintOptions which allows clients to decide what they would like to use for the delimiter rather than trying to modify the default delimiter.

If it seems OK, I will proceed with #37978 and #37981 for now to see if the community feels comfortable with this approach. If these changes end up getting merged, then it may make sense to mark #30951 as closed since it kind of feels like #37981 partially address it (although, admittedly, it isn't exactly the same, since it doesn't change the default display to include a space after the comma).

[rok]

@kevingurney Thank you for the extensive and thoughtful explanation of your proposal! :D
I think it's best you proceed with your proposal and indeed close out #30951 once your work is merged.