Draft of book's core package (Zipfs Law)

[lwjohnst86]

No need to have too much done on it, mostly to get an idea of what it will look like. For now, a repo in merely-useful is fine.

[cwickham]

I have a draft up at: https://github.com/merely-useful/zipfs

At this point it's probably useful to:

• review the vignette, which runs through all the functionality, and
• glance over the body of the functions in R/

Some notes:

• In terms of interface and implementation I assumed a tidyverse style approach, but at the same time avoiding having to deal with any tidy evaluation. This means the body of functions use things like stringr, tibble, ggplot2 and the pipe.

• book_meta() didn't exist in the Python version, but it basically implements some of the things explored in the Shell chapters of py-rse. It might be a good candidate for introducing functions (and testing), since the body is relatively straight-forward.

• My word counts for Dracula don't quite match those in py-rse...any thoughts on why?

• I thought a little about how this relates to the realistic task of applying this to many books in a directory. For now, I've put the books from py-rse in inst/extdata and some incomplete notes in dev/many-books.Rmd.

[cwickham]

Realized discussion is probably better handled in the draft repo. I've opened an issue at: merely-useful/zipfs#1