implement report()

[mrcaseb]

closes #274

added the function report() that should help us debug problems by outputting version information

[mrcaseb]

This also includes a nice 3 column implementation of dependencies @tanho63

nflfastR::report("nflfastR", TRUE)
#> -- System Info -----------------------------------------------------------------
#> * R version 4.1.1 (2021-08-10)
#> * OS: x86_64-w64-mingw32/x64 (64-bit)
#> -- Packages --------------------------------------------------------------------
#> * nflfastR (4.2.0.9006)
#> -- Dependencies ----------------------------------------------------------------
#> * cli        (3.0.1)     * codetools   (0.2-18)    * cpp11      (0.3.1) 
#> * crayon     (1.4.1)     * curl        (4.3.2)     * data.table (1.14.0)
#> * digest     (0.6.27)    * dplyr       (1.0.7)     * ellipsis   (0.3.2) 
#> * fansi      (0.5.0)     * fastrmodels (1.0.2)     * furrr      (0.2.3) 
#> * future     (1.22.1)    * generics    (0.1.0)     * globals    (0.14.0)
#> * glue       (1.4.2)     * janitor     (2.1.0)     * jsonlite   (1.7.2) 
#> * lattice    (0.20-44)   * lifecycle   (1.0.0)     * listenv    (0.8.0) 
#> * lubridate  (1.7.10)    * magrittr    (2.0.1)     * Matrix     (1.3-4) 
#> * mgcv       (1.8-36)    * nlme        (3.1-152)   * parallelly (1.28.1)
#> * pillar     (1.6.2)     * pkgconfig   (2.0.3)     * progressr  (0.8.0) 
#> * purrr      (0.3.4)     * R6          (2.5.1)     * Rcpp       (1.0.7) 
#> * rlang      (0.4.11)    * snakecase   (0.11.0)    * stringi    (1.7.4) 
#> * stringr    (1.4.0)     * tibble      (3.1.4)     * tidyr      (1.1.3) 
#> * tidyselect (1.1.1)     * utf8        (1.2.2)     * vctrs      (0.3.8) 
#> * xgboost    (1.4.1.1)   * codetools   (0.2-18)    * cpp11      (0.3.1)

^{Created on 2021-09-16 by the reprex package (v2.0.1)}

[guga31bb]

If we use the new pipe we have to increment the R version requirement in DESCRIPTION, right? I would be inclined to not do this (I can't even test this bc I don't have this version of R, lol)

[mrcaseb]

If we use the new pipe we have to increment the R version requirement in DESCRIPTION, right? I would be inclined to not do this (I can't even test this bc I don't have this version of R, lol)

Good catch! That was a mistake. Will fix

[tanho63]

I'm not opposed to bringing in cli at nflreadr level if it makes more sense for this to live there, does it make more sense?

[guga31bb]

Ok looks good to me!

[mrcaseb]

I'm not opposed to bringing in cli at nflreadr level if it makes more sense for this to live there, does it make more sense?

I wanted to test the function with cli and found two special cases that need to be caught haha. Adding cli to nflreadr would also add glue and I'd like to keep it light so I would put report into nflfastR tbh

nflfastR::report("cli", TRUE)
#> -- System Info -----------------------------------------------------------------
#> * R version 4.1.1 (2021-08-10)
#> * OS: x86_64-w64-mingw32/x64 (64-bit)
#> -- Packages --------------------------------------------------------------------
#> * cli (3.0.1)
#> -- Dependencies ----------------------------------------------------------------
#> * glue (1.4.2)

[tanho63]

trouble to me is it becomes annoying to re-export a report function if we're debugging anything other than nflfastR?

[mrcaseb]

trouble to me is it becomes annoying to re-export a report function if we're debugging anything other than nflfastR?

I am not sure what that means tbh. I assume most users that are asking us nflverse related questions will have installed nflfastR anyways. And if that's not the case they just need to install it an run the function. Sounds like a fair price for not adding another two deps to nflreadr. But that is not a strong opinion.

[mrcaseb]

Going to merge this now. We can still move the function in another package if we feel like it is necessary. It's not a function that is a typical use case in any user script so moving it is not a big deal at all