Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add potential sequence() lints to seq_linter() #2618

Open
wants to merge 11 commits into
base: main
Choose a base branch
from
Open

Add potential sequence() lints to seq_linter() #2618

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
32049af
Draft sequence_linter linter
Bisaloo Jun 7, 2024
332f861
Fold sequence_linter() into seq_linter()
Bisaloo Jun 21, 2024
405644e
Add tests
Bisaloo Jun 21, 2024
999034c
Add NEWS bullet
Bisaloo Jun 21, 2024
75e56a4
Switch to xml_find_function_calls
Bisaloo Jun 22, 2024
69b4ad2
Change element order in xpath
Bisaloo Jun 22, 2024
396a239
Support lapply(x, seq)
Bisaloo Jun 22, 2024
231abff
Simplify xpath
Bisaloo Jun 22, 2024
cfc17ee
Do not lint seq() calls with extra arguments
Bisaloo Jun 22, 2024
072ef13
Edit false positive example
Bisaloo Jun 22, 2024
2a8d243
Add more info about new lints in NEWS
Bisaloo Jun 22, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions NEWS.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,7 @@

## New and improved features

* `seq_linter()` now includes lints to inform about missed opportunities to use the `sequence()` base R function (#2618, @Bisaloo)
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think sequence() is not a very well-known function (<500 calls on CRAN vs more than 80K for seq()), here is a good opportunity to quickly introduce it:

"to use the sequence() function, e.g. unlist(lapply(ints, seq))"

* More helpful errors for invalid configs (#2253, @MichaelChirico).
* `library_call_linter()` is extended
+ to encourage all packages to be attached with `library(symbol)`, not `library("symbol", character.only = TRUE)` or "vectorized" approaches looping over package names (part of #884, @MichaelChirico).
Expand Down
31 changes: 30 additions & 1 deletion R/seq_linter.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,6 +26,11 @@
#' linters = seq_linter()
#' )
#'
#' lint(
#' text = "unlist(lapply(x, seq_len))",
#' linters = seq_linter()
#' )
#'
#' # okay
#' lint(
#' text = "seq_along(x)",
Expand All @@ -42,6 +47,11 @@
#' linters = seq_linter()
#' )
#'
#' lint(
#' text = "sapply(x, seq_len)",
IndrajeetPatil marked this conversation as resolved.
Show resolved Hide resolved
#' linters = seq_linter()
#' )
#'
#' @evalRd rd_tags("seq_linter")
#' @seealso [linters] for a complete list of linters available in lintr.
#' @export
Expand All @@ -66,6 +76,15 @@ seq_linter <- function() {
]
")

map_funcs <- xp_text_in_table(c("sapply", "lapply", "map"))
sequence_xpath <- glue("
//SYMBOL_FUNCTION_CALL[ { map_funcs } ]
/parent::expr/parent::expr[
preceding-sibling::expr/SYMBOL_FUNCTION_CALL[text() = 'unlist']
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: I would try and keep the logic in one direction on the AST: check sapply, then check seq_len, then check unlist, roughly

//SYMBOL_FUNCTION_CALL/parent::expr[following-sibling::expr/SYMBOL]/parent::expr/parent::expr[expr/SYMBOL_FUNCTION_CALL]

Also, please switch to use the new find_function_calls approach, which is appropriate here, e.g.

lintr/R/condition_call_linter.R

Line 83 in 6ed78c2

xml_calls <- source_expression$xml_find_function_calls(c("stop", "warning"))

and expr/SYMBOL[text() = 'seq_len']
]"
)

## The actual order of the nodes is document order
## In practice we need to handle length(x):1
get_fun <- function(expr, n) {
Expand Down Expand Up @@ -113,6 +132,16 @@ seq_linter <- function() {
)
)

xml_nodes_to_lints(badx, source_expression, lint_message, type = "warning")
seq_lints <- xml_nodes_to_lints(badx, source_expression, lint_message, type = "warning")

potential_sequence_calls <- xml_find_all(xml, sequence_xpath)
sequence_lints <- xml_nodes_to_lints(
potential_sequence_calls,
source_expression,
"Use sequence() to generate a concatenated sequence of seq_len().",
type = "warning"
)

c(seq_lints, sequence_lints)
})
}
10 changes: 10 additions & 0 deletions man/seq_linter.Rd
View file
Open in desktop

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

30 changes: 30 additions & 0 deletions tests/testthat/test-seq_linter.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -123,6 +123,24 @@ test_that("reverse seq is ok", {
)
})

test_that("finds potential sequence() replacements", {
linter <- seq_linter()
lint_msg <- rex::rex("Use sequence()")

expect_lint(
"unlist(lapply(x, seq_len))",
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The lint now only covers unlist(lapply(x, seq_len)).

What about

  • unlist(lapply(x, seq))
  • unlist(lapply(x, seq, by = 2)), and other variants equivalent to using the from= and/or by= arguments of sequence()?
  • I don't think it's worth trying to find examples like lapply(x, \(xi) seq(xi)) which would be covered by unnecessary_lambda_linter(). But are there any examples that do require looking into a lambda to replicate a sequence()-able call?

(it would be nice to get a sense of how many hits these are getting before investing too much time -- and also feel free to earmark some of this as a TODO in a follow-up issue, though lapply(x,seq) is trivial enough to be done here)

Copy link
Contributor Author

@Bisaloo Bisaloo Jun 22, 2024 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

unlist(lapply(x, seq, by = 2)), and other variants equivalent to using the from= and/or by= arguments of sequence()?

I don't think there is a direct equivalence:

  • from has a different behaviour in seq() and sequence()

    unlist(lapply(1:5, seq, from = 2))
#>  [1] 2 1 2 2 3 2 3 4 2 3 4 5

sequence(1:5, from = 2)
#>  [1] 2 2 3 2 3 4 2 3 4 5 2 3 4 5 6

    Created on 2024-06-22 with reprex v2.1.0

  • As far as I can tell, by= cannot be used on its own without from=, which brings us back to the previous case

    seq(20, by = 2)
#> Error in seq.default(20, by = 2): wrong sign in 'by' argument

    Created on 2024-06-22 with reprex v2.1.0

I have edited the xpath to make sure this type of call doesn't lint.

lint_msg,
linter
)

# Even for prefixed purrr:: calls
expect_lint(
"unlist(purrr::map(x, seq_len))",
lint_msg,
linter
)
})

test_that("Message vectorization works for multiple lints", {
linter <- seq_linter()

Expand Down Expand Up @@ -173,6 +191,18 @@ test_that("Message vectorization works for multiple lints", {
),
linter
)

expect_lint(
trim_some("{
1:NROW(x)
unlist(lapply(y, seq_len))
}"),
list(
list(rex::rex("seq_len(NROW(...))", anything, "1:NROW(...)"), line_number = 2L),
list(rex::rex("sequence()"), line_number = 3L)
),
linter
)
})

test_that("Message recommends rev() correctly", {
Expand Down
Loading