follow coding style guidelines enforced by lintr

[mschilli87]

The following lintr linters currently cannot be enabled without breaking the tests:

• [x] infix_spaces_linter (0d0bfe8dac2baf9d1a820f75213fc92272e72584)
• camel_case_linter
• [x] line_length_linter (994540090829d993ef43a48e457c5013c7f10b80)
• [x] single_quotes_linter (c0bbfae1fac6ed824715f23be47b036d4c3f2614)
• multiple_dots_linter
• [x] assignment_linter (db3f2d973c4e960a8422ed7f18d674d7be92e20f)
• [x] commas_linter (9e2b16a9f754aef39f664e84254fffe63e0aab4d)
• [x] commented_code_linter (b87b4539d89391531531c35eab0a5d6c568fc26d)
• object_usage_linter (bf41f5bd3c3a91ad0cdb2d50e3c9bdf7ab3685f0)
• [x] spaces_left_parentheses_linter (2cab2062778895c6c514450b61f01ded88c0e8b1)
• [x] absolute_paths_linter (0f9b7b9b427d8984b919c325cc102cffedbb06eb)
• [x] object_length_linter (b3df7acc0da4a87101244cb3b5dc3e6ff5064aa5)
• [x] open_curly_linter (cbdf6d7c3ca6d2c0a87c56b38dece718696e3caa)
• [x] closed_curly_linter (78cbbc9cc82eb07fb2d1299b45e5eb9362b31132)
• [x] no_tab_linter (5d450b30570eaa51227ce7543849fcb3cbd234fd)
• [x] trailing_blank_lines_linter (53d0cbc1c4a8f6de2bd39fd6cb4efd4d4fa78bc7)

Unless we agree to skip any of those for some specifc reason, the code should be adjusted accordingly and the linters be part of the test suite.

[mschilli87]

I think object_usage_linter will be hard to satisfy (see http://stackoverflow.com/q/9439256/2451238). Even though I think it makes sense (what happens if your data.frame doesn't have a column of that name?), and I'd be willing to replace aes by aes_string (see http://stackoverflow.com/a/12429344/2451238) without really solving the underlying 'problem', I have no clue how to 'fix' this for data.tables.

@retaj: Any thoughts on this?

---

edit: See bf41f5bd3c3a91ad0cdb2d50e3c9bdf7ab3685f0 to see what object_usage_linter complains about.

[mschilli87]

I think we definitely should satisfy absolute_paths_linter, but I don't know what part of code is just some old leftover test and how to adjust the parts that cannot be deleted (yet).

@retaj: Can you have a look at 78551cb81730c98939dbd46e73eb7870b01cca6c and point me to how to work around the absolute paths?

[retaj]

test_load_data.R should be excluded from linting, it's just my debugging scrapbook...

object_usage_linter should be ignored (just like a bunch of R CMD check notes that get raised when it encounters ggplot2, data.table or plyr code). I know there are workarounds, but I don't think we should obfuscate our code to satisfy outdated checks and styling guidelines.

[mschilli87]

• multiple_dots_lintr chokes on names like circ.gr.reduced or collapse.to.other:

  Words within variable and function names should be separated by '_' rather than '.'

  • camel_case_linter (obviously) has a problem with things like colData, queryHits and rowRanges that would need conversion of some x$y to x[["y"]] plus explicit usage of :: to explicetely exclude foreign function names from linting: Variable and function names should be all lowercase.

My guess is that we'd rather keep these linters disabled instead of re-naming functions/variables and/or go all [[/:: (at least for now).

@retaj: Do you agree?

If so, I will remove them from the todo list of this issue and once you have resolved #30, I can finish off this one for good.

[retaj]

I think both dots and underscores in variable names are fine. Guess each style guide has a different preference (e.g. google style guide is strictly against underscores in R function names), but most people are comparably good at reading variable.names, variable_names and variableNames.

I would keep these two off...