sanitize_token() doesn't allow lowercase characters

[jmbarbone]

Adding a small change in sanitize_token() with toupper() could save some work for people with API tokens stored with lowercase character data.

library(REDCapR)
sanitize_token("12345678901234567890123456ABCDEF")
#> [1] "12345678901234567890123456ABCDEF"
sanitize_token("12345678901234567890123456abcdef")
#> Error in sanitize_token("12345678901234567890123456abcdef"): The token is not a valid 32-character hexademical value.
packageVersion("REDCapR")
#> [1] '1.0.0'

^{Created on 2021-07-22 by the reprex package (v2.0.0)}

[wibeasley]

Thanks @jmbarbone. I restricted to uppercase intentionally, because REDCap's token generating code wasn't producing any lowercase letters. I assume it generated lowercase letters for you? I guess their code changed.

If so, do you mind removing the call to toupper(), and instead loosen the regex so it accommodates lowercase letters from a to z? The pattern would then be defined like this (notice the extra "a-f")

  pattern <- "^([0-9A-Fa-f]{32})(?:\\n)?$"

[jmbarbone]

Yes, @wibeasley, all the API tokens I've seen use lowercase characters.

The reason I used toupper() instead of adjusting the regex was to make sure that the outputs would be the same regardless of the original case of token in case there were other checks that demanded upper case tokens. But from local testing it seems that adjusting just the pattern is safe.

[wibeasley]

I'm glad you figured it out and told me instead of turning off the validation.