Merge a3bceb7 into 984a399

Author: MichaelChirico

DESCRIPTION

@@ -156,6 +156,7 @@ Collate:
     'regex_subset_linter.R'
     'repeat_linter.R'
     'routine_registration_linter.R'
+    'sample_int_linter.R'
     'scalar_in_linter.R'
     'semicolon_linter.R'
     'seq_linter.R'

NAMESPACE

@@ -119,6 +119,7 @@ export(redundant_ifelse_linter)
 export(regex_subset_linter)
 export(repeat_linter)
 export(routine_registration_linter)
+export(sample_int_linter)
 export(sarif_output)
 export(scalar_in_linter)
 export(semicolon_linter)

NEWS.md

@@ -6,6 +6,7 @@
 
 ### New linters
 
+* `sample_int_linter()` for encouraging `sample.int(n, ...)` over equivalents like `sample(1:n, ...)` (part of #884, @MichaelChirico).
 * `stopifnot_all_linter()` discourages tests with `all()` like `stopifnot(all(x > 0))`; `stopifnot()` runs `all()` itself, and uses a better error message (part of #884, @MichaelChirico).
 * `comparison_negation_linter()` for discouraging negated comparisons when a direct negation is preferable, e.g. `!(x == y)` could be `x != y` (part of #884, @MichaelChirico).
 * `terminal_close_linter()` for discouraging using `close()` to end functions (part of #884, @MichaelChirico). Such usages are not robust to errors, where `close()` will not be run as intended. Put `close()` in an `on.exit()` hook, or use {withr} to manage connections with proper cleanup.

R/sample_int_linter.R

@@ -0,0 +1,60 @@
+#' Require usage of sample.int(n, m, ...) over sample(1:n, m, ...)
+#'
+#' [sample.int()] is preferable to `sample()` for the case of sampling numbers
+#'   between 1 and `n`. `sample` calls `sample.int()` "under the hood".
+#'
+#' @evalRd rd_tags("sample_int_linter")
+#' @seealso [linters] for a complete list of linters available in lintr.
+#' @export
+sample_int_linter <- function() {
+  # looking for anything like sample(1: that doesn't come after a $ extraction
+  # exclude TRUE/FALSE for sample(replace = TRUE, ...) usage. better
+  #   would be match.arg() but this also works.
+  xpath <- glue("
+  //SYMBOL_FUNCTION_CALL[text() = 'sample']
+    /parent::expr[not(OP-DOLLAR or OP-AT)]
+    /following-sibling::expr[1][
+      (
+        expr[1]/NUM_CONST[text() = '1' or text() = '1L']
+        and OP-COLON
+      )
+      or expr/SYMBOL_FUNCTION_CALL[text() = 'seq_len']
+      or (
+        expr/SYMBOL_FUNCTION_CALL[text() = 'seq']
+        and (
+          count(expr) = 2
+          or (
+            expr[2]/NUM_CONST[text() = '1' or text() = '1L']
+            and not(SYMBOL_SUB[
+              text() = 'by'
+              and not(following-sibling::expr[1]/NUM_CONST[text() = '1' or text() = '1L'])
+            ])
+          )
+        )
+      )
+      or NUM_CONST[not(text() = 'TRUE' or text() = 'FALSE')]
+    ]
+    /parent::expr
+  ")
+
+  Linter(function(source_expression) {
+    if (!is_lint_level(source_expression, "expression")) {
+      return(list())
+    }
+
+    xml <- source_expression$xml_parsed_content
+
+    bad_expr <- xml_find_all(xml, xpath)
+    first_call <- xp_call_name(bad_expr, depth = 2L)
+    original <- sprintf("%s(n)", first_call)
+    original[!is.na(xml_find_first(bad_expr, "expr[2]/OP-COLON"))] <- "1:n"
+    original[!is.na(xml_find_first(bad_expr, "expr[2]/NUM_CONST"))] <- "n"
+
+    xml_nodes_to_lints(
+      bad_expr,
+      source_expression = source_expression,
+      lint_message = glue("sample.int(n, m, ...) is preferable to sample({original}, m, ...)."),
+      type = "warning"
+    )
+  })
+}

inst/lintr/linters.csv

@@ -76,6 +76,7 @@ redundant_ifelse_linter,best_practices efficiency consistency configurable
 regex_subset_linter,best_practices efficiency
 repeat_linter,style readability
 routine_registration_linter,best_practices efficiency robustness
+sample_int_linter,efficiency readability robustness
 scalar_in_linter,readability consistency best_practices efficiency
 semicolon_linter,style readability default configurable
 semicolon_terminator_linter,style readability deprecated configurable