Apply coding conventions [WIP]

[se-wo]

I started applying coding conventions to the code to improve readability and maintainability. I used Tidyverse style as the basis. I added a basic configuration for Lintr and configured a corresponding Github action. In addition I applied Styler to fix all automatically fixable issues.

Nevertheless there are still more than enough style related issues. Code style is to some extend art, so it´s difficult to argue whether or not tidyverse style is appealing. Style policy modifications can be made in .lintr. E.g.

linters: linters_with_defaults(
    line_length_linter(120)
    )
encoding: "UTF-8"

Current linter findings (Details)

Linter	# Occurrences	Comment
commented_code_linter	72	:x: CRAN package should not contain commented out code
cyclocomp_linter	1	:warning: No showstopper but could improve maintainability
line_length_linter	100	:warning: Maybe extend to 120 characters and fix rest?
object_name_linter	259	:information_source: Would be quite a lot of work
object_usage_linter	20	:x: Unused variables should be cleaned up
seq_linter	18	:warning: cause of errors if nrow/ncol is 0
T_and_F_symbol_linter	33	:information_source: But easy to fix
vector_logic_linter	2	:warning: Is using a single \| really intentional?
Total	505

[se-wo]

Open for discussion on how configure the linter rules

[RuedigerForster]

@se-wo: Your proposal is fine. Line length should not exceed 100 characters; I wlll try and keep it around 80. I will also fix commented_code and line_length, when the code is close to final. You don't need to do this. |: (simple or) was intentional: 2 logical expressions should be evaluated at development time. Can be replaced by || in final version. Can we have a short discussion about object_usage and object_name? My current scheme is: Classes and objects: UpperCamelCase methods: lowerCamelCase internal functions and their arguments: lower_snake_case exported functions and their arguments: lower.dot.case

[RuedigerForster]

@se-wo: Your proposal is fine. Line length should not exceed 100 characters; I wlll try and keep it around 80. I will also fix commented_code and line_length, when the code is close to final. You don't need to do this. |: (simple or) was intentional: 2 logical expressions should be evaluated at development time. Can be replaced by || in final version. Can we have a short discussion about object_usage and object_name? My current scheme is: Classes and objects: UpperCamelCase methods: lowerCamelCase internal functions and their arguments: lower_snake_case exported functions and their arguments: lower.dot.case

lintr is quite dumb. I did the best I could do. Left the dot.case for items in the S4 classes. Other names changed according to: https://cran.r-project.org/web/packages/easystats/vignettes/conventions.html

lintr does not recognize .not_exported_function, and variable names which are hidden in lists or function calls.

Commented code still there to some extent. Will delete later.

[RuedigerForster]

Added lintr workflow to source.