ofToInt, ofToFloat, etc conflate error status with valid values

The ofToX family of functions return a value of 0 (or 0.0) on error even though that is also a valid conversion value. Also, for ofToInt and ofToInt64 functions, the documentation is not accurate since an out of range string will produce the min/max integer value as the error code, not zero.

For example, the following code:

printf("%d\n", ofToInt("0"));
printf("%d\n", ofToInt(""));
printf("%d\n", ofToInt("99999999999"));

outputs the following:

0
0
2147483647

Line 1 and line 2 return the same value, even though line 1 succeeds and line 2 fails. Line 3 fails but returns max int instead of zero.

For a solution, I was thinking of something like the following, which would avoid breaking the current API:

template<typename T>
T ofTo(const std::string & str, bool *fail = nullptr){
	T x;
	std::istringstream cur(str);
	cur >> x;
        if (fail) {
            *fail = iss.fail();
        }
	return x;
}

The return value would not change, but the optional "fail" flag would be set to indicate a conversion error. If this is acceptable, I would be happy to create a PR.

Start by locating the ofToX family definitions and its documentation, then inspect existing tests or call sites for conversion errors. Verify how zero and out-of-range results are currently reported, and define a compatible way to distinguish successful zero conversions from failures while correcting the documented behavior.