Clean up some uses of String::substr

[AThousandShips]

Cases where the end position is either equvalent to the default or past the end of the string.

Cases:

• X.substr(Y, X.length()) (which is always equivalent or beyond the end)
• X.substr(Y, X.length() - Y), X.substr(Y, X.size() - (Y + 1)) (which are always equivalent)
• X.substr(Y, X.size()) or X.substr(Y, X.size() - Y) (which is always beyond the end)
• X.substr(Y, -1) (which is the default)

For context, X.substr(Y, -1) is equivalent to X.substr(Y, X.length() - Y), and if the second argument is greater than X.length() - Y it is clamped, so equivalent to the default again. Note that X.size() is always X.length() + 1 unless X.length() == 0 (in which case substr returns empty)

This communicates this intent more clearly, and avoids unnecessary operations, and avoids errors due to copying the code or minor changes down the line.

Didn't try to identify cases equivalent to this where the second argument was a variable.

[Ivorforce]

I skimmed all changes; looks correct to me. I agree it's cleaner without an explicit second argument.

I would recommend waiting to merge until 4.5 due to the non zero risk of simply having gotten one of those wrong.

[AThousandShips]

Absolutely for no sooner than 4.5, went over them with a few specific patterns (including some negations like X.substr(A + B, X.size() - A - B) and similar which were a bit harder to detect

Can split into individual categories for ease of review (i.e. the trivial case of X.substr(Y, X.length()), but shouldn't be difficult to work through the validity of them

[bruvzg]

Looks good.