[R] Consider ways to make arrow.skip_nul option more user-friendly

[asfimport]

In Arrow 3.0.0, the arrow.skip_nul option effectively defaults to FALSE for consistency with base::readLines and base::scan.

If the user keeps this default option value, then conversion of string data containing embedded nuls causes an error with a message like:

embedded nul in string: '\0' 

If the user sets the option to TRUE, then no error occurs, but this warning is issued:

Stripping '\0' (nul) from character vector 

Consider whether we should:

1. Keep this all as it is
2. Change the default option value to TRUE
3. Keep the default option value as it is, but catch the error and re-throw it with a more actionable message that tells the user how to set the option

Reporter: Ian Cook / @ianmcook
Assignee: Neal Richardson / @nealrichardson

Related issues:

• [R] Arrow to R fails with embedded nuls in strings (relates to)
• [R] Regression from 2.0.0 -> 3.0.0: Null character in string prevents dataset from loading (is related to)

PRs and other links:

• GitHub Pull Request #9899

_{Note: This issue was originally created as ARROW-11478. Please see the migration documentation for further details.}

[asfimport]

Ian Cook / @ianmcook:
I think I'm in favor of option 3, assuming it's feasible (which I think it should be).

[asfimport]

Jonathan Keane / @jonkeane:
Are there big downsides to option 2?

[asfimport]

Neal Richardson / @nealrichardson:
The skip_nul path is going to be significantly slower; see the code here: https://github.com/apache/arrow/blob/master/r/src/array_to_vector.cpp#L353-L391

We can do option 3; probably that has to be done on the R side.

[asfimport]

Neal Richardson / @nealrichardson:
Issue resolved by pull request 9899
#9899