Use correct regexp lowercase \z terminator

[johnnyshields]

In VALID_DECIMAL128_STRING_REGEX, a recent commit used \Z (uppercase). Probably \z (lowercase) is the correct one. The current regex will allow "21.23\n" as a valid value.

$ Matches the end of a line.

\z Matches the end of the string.

\Z Matches the end of the string unless the string ends with a ``\n'', in which case it matches just before the ``\n''.

https://ruby-doc.com/docs/ProgrammingRuby/html/language.html#UJ

[jamis]

Hey @johnnyshields -- thanks for the PR. Is the concern that a string with a newline matches, and you believe it shouldn't?

Looking at the original regex (prior to that commit you mentioned), it used /^...$/ anchors, and also matches the newline:

"21.23\n" =~ /^[\-\+]?(\d+(\.\d*)?|\.\d+)(E[\-\+]?\d+)?$/i # => 0

The \Z anchor preserves this behavior, whereas (as you noted) the lowercase \z does not. It should be noted that Float("21.23\n") also succeeds, so allowing a trailing newline has some precedent in Ruby itself.

If I'm misunderstanding your concern, please let me know. Otherwise, I'm inclined to let the current implementation stand.

It should be noted in passing that the regex was changed at the recommendation of the CodeQL static analyzer, which offered \Z as the recommended alternative to $. I had no compelling reason to prefer one over the other, so opted to appease CodeQL.

[johnnyshields]

If being permissive of the trailing newline is intentional, then fine to keep it. (In general \z is preferable to \Z; in the vast majority of regexp use cases one does not intend to allow trailing newlines.)