match.arg is not behaving as I thought it did

[CeresBarros]

match.arg won't error if arg contains entries that are not found in choices -- it simply excludes them. It also does a partial which may lead to selecting the wrong options (especially for variable names if the user partially writes them right.

This is not the type of behaviour we want most times so I replaced many of the match.arg with setdiff-type of checks.

My local checks are failing to build one of the vignettes, but I doubt that this is related to my changes - keeping this as a draft to diagnose.

Also, please note that I've added 00 to the development version numbering (it should be 9000, 9001, 9002, ..., not 90, 91, 92, ...)

[CeresBarros]

climr_workflow_int.Rmd is failing to knit with RStudio build tools, but not if I knit it manually -- so possibly okay?
Marking PR as ready to review

[CeresBarros]

uhmm i see the same vignette build failing with a different error now on GHA:

Error: Error in proc$get_built_file() : Build process failed

[CeresBarros]

other fixes have been added to this PR:

• system.file doesn't find the file if devtools::load_all was used to load climr
• removal of nested function definition
• added missing imports and global variable definitions
• streamlining of code in plot_timeSeries_input and related internal functions

[CeresBarros]

all tests seem to be passing now, except code coverage calculation -- not sure what's going on, but not serious.

@cmahony before merging, could you please check that the plot_timeSeries_input function is behaving as intended? I did some important code changes (streamlining, readability, moved internal function definitions to "outside") and things look good to me (the code in te @examples) but perhaps having your eyes on it would be better. Thanks :)

[CeresBarros]

approval on plot_timeSeries function (not plot_timeSeries_input, my mistake on an earlier comment) by @cmahony -- all code checks are passing, so I'm merging.