Skip to contents

Attribute checks - Names

has_names_valid.Rd

Test whether x has names with greater flexibility than rlang::is_named(), handling NA, empty, duplicate, and invalid names, as well as empty vectors.

Usage

has_names_valid(
  x,
  na = FALSE,
  empty = FALSE,
  dups = FALSE,
  invalid = TRUE,
  zero_len = TRUE,
  how = "names"
)

are_names_valid(
  x,
  na = FALSE,
  empty = FALSE,
  dups = FALSE,
  invalid = TRUE,
  how = "names"
)

Arguments

x

[collection(), any] For are_*(), any collection to test; for is_*(), any object to test.

na, empty, dups, invalid

[TRUE | FALSE] What to return for NA, empty, duplicate, or invalid names.

zero_len

[TRUE | FALSE] What to return for an an empty x.

how

["names" | "x" | "attr" | "names2"] How to extract names: "names" for names(); "x" to use x itself; "attr" for attr(x, "names"); or "names2" for rlang::names2().

Value

[TRUE | FALSE] The scalar result of the test.

Details

The default argument values guarantee that x can be safely iterated by its names.

A non_collection is an object that is not an atomic vector, list, pairlist, expression, or environment. That is, it returns FALSE for is_collection().

The how methods differ only for environments, with "names" returning names and "attr" returning NULL.

invalid = FALSE checks if the names don't change after being passed to make.names().

Examples

has_names_valid(mtcars) #> TRUE
#> [1] TRUE

# NA, empty, and duplicate names return FALSE by default:
has_names_valid(rlang::set_names(1:3, c("a", "b", NA))) #> FALSE
#> [1] FALSE
has_names_valid(rlang::set_names(1:3, c("a", "b", NA)), na = TRUE) #> TRUE
#> [1] TRUE

has_names_valid(rlang::set_names(1:3, c("a", "", "b"))) #> FALSE
#> [1] FALSE
has_names_valid(rlang::set_names(1:3, c("a", "a", "a"))) #> FALSE
#> [1] FALSE

# Invalid names and zero-length x's return TRUE by default:
has_names_valid(rlang::set_names(1:3, c("_bad", "b", "c"))) #> TRUE
#> [1] TRUE
has_names_valid(integer(0)) #> TRUE
#> [1] TRUE

# Use `how` to control how names are extracted:
has_names_valid(rlang::base_env(), how = "attr") #> FALSE
#> [1] FALSE
# (envs have no names attribute)
has_names_valid(c("a", "b", "c"), how = "x") #> TRUE (x itself was tested)
#> [1] TRUE

# Vectorized test:
are_names_valid(c("a", "b", "b", NA, ""), how = "x")
#> [1]  TRUE FALSE FALSE FALSE FALSE
#> c(TRUE, FALSE, FALSE, FALSE, FALSE)