how to validate readLines() input as numeric vs. character in R

I am coercing a input from readLines() which I understand will take any input, character or numeric, and capture it as a character value. I then want to convert the input to a numeric value using as.numeric().

I then want to use a validation test such as is.numeric() to check that user has input 5 vs a character string as "five" with the assertthat package.

However when I coerce a character vector (eg. "five") to as.numeric() it will produce a NA output. when you test an NA output with is.numeric() it will produce a TRUE defeating the whole purpose

How can I validate that a user input a numeric value given that readLines() converts everything to a character vector?

#both of the below produce TRUE 
  is.numeric(as.numeric("five"))
  is.numeric(as.numeric(5))

Solution

I believe you can use !is.na to test the result of as.numeric. If it failed to convert to valid numeral values, as.numeric returns NA.

!is.na(as.numeric("five"))
#> Warning: NAs introduced by coercion
#> [1] FALSE
!is.na(as.numeric("5"))
#> [1] TRUE
!is.na(as.numeric(5))
#> [1] TRUE

Or use the regex ^[0-9]+$ to see if the whole input is made up entirely of integers (for testing integers only, not for testing floating point numbers).

grepl("^[0-9]+$", "five")
#> [1] FALSE
grepl("^[0-9]+$", "fi55ve")
#> [1] FALSE
grepl("^[0-9]+$", "356236345")
#> [1] TRUE
grepl("^[0-9]+$", 5235)
#> [1] TRUE

^{Created on 2023-09-15 with reprex v2.0.2}