Closed petergoldstein closed 2 years ago
@gettalong Yeah, I was aware of the slight behavior change but I don't think it's an issue because:
options[:content]
was nil
- options[:content]
was being set to the empty string.@pointlessone This does override #103. If we merge this PR (which is my preference), we can close #103 . Given that @schlick - the original author of #103 - gave this a thumbs up, I assume that he's good with that approach as well.
This is a combination of #103 and #119 with tests for both cases. Thanks to @schlick and @tetsuya-ogawa for the original PRs.
@pointlessone Since the original PRs overlap, and tests were only included in one of them, I think this is a good combination PR. Missing tests were added.
It also makes it easy to add future types to be stringified and, should it be desired, to change the behavior to stringify anything but
Prawn::Table
,Prawn::Table::Cell
, or Array that responds to ato_s
@gettalong any comments are welcome as well.