Change reader signature

[Hugovdberg]

Major change of the interface, this is a pull request against the release-1.0.0 branch so we can clean up the current 0.8.2 release and finalise it into the 0.8.3 release. This feature needs some thorough testing and should not be released in 0.8.3 in my opinion.

Types of change

• [ ] Bugfix (non-breaking change which fixes an issue)
• [x] New feature (non-breaking change which adds functionality)
• [x] Breaking change (fix or feature that would cause existing functionality to not work as expected)
• [ ] Other (documentation etc)

Pull request checklist

• [x] Add functionality
• [x] Add tests
• [x] Update documentation in man
• [ ] Update website documentation

---

Proposed changes

• I created a new version of .add.extension that stores the passed in reader with the matching extension. Unlike previous versions an actual reference to the function is stored instead of a character string that is evaluated when loading the data. To be able to store and display the reader nicely in a data.frame I created a new R6 class that stores the function and the name. I then changed list.data to actually load it into a tibble because tibbles handle non standard columns better. Also load.project actually calls the reader function inside the R6 object instead of looking up the function by name.
• I changed the signature of all the reader functions and add them in the same file to the dispatch table by calling .add.extension. This way the reader and the code adding it to the list of available readers is more or less self contained. The preinstalled.readers list is now a function that can actually be used to inspect which readers are available. Perhaps this function should be renamed available.readers.
• Because R6 and tibble are now actual dependencies of the package I added them to the Imports section in DESCRIPTION
• Because xls.reader and xlsx.reader are actually the same reader since we migrated to readxl I removed the xlsx.reader and added both extensions to the xls.reader.

[KentonWhite]

@Hugovdberg Is this ready for review or still working on it?

[Hugovdberg]

I think it is ready for review, I think further optimization and cleaning is possible, but I prefer to do it after you reviewed it. Keep in mind I propose to only introduce this in 1.0. Also, the dependency on tibble and R6 should be too much of a problem because they are both dependencies of dplyr, which will be introduced with the tidyverse anyway. R6 classes have a nice lightweight profile, and act more like classical OOP where data and methods are kept together. I first tried to implement it as an S3 class but ended up needlessly cluttering the interface with a lot of auxiliary methods that R6 just doesn't need.