"Number of a string VS Number of a number" AKA Conversion problems, help!

programming flash
1

Hello all,
i have a big problem, and i’d like some advice from you on how to solve it.

Inside my code i have some variables declared as Numbers, and i get some others from an external source; these variables can either be Numbers or String representation of Numbers.

Now, i have to implement some equality controls on these variables, and since i don’t know in that moment which type are them of, i’m converting all of them to Numbers (i don’t need and neither want to compare strings) via the Number() function which accepts either Numbers or Strings as an input.

These controls can compare 2 single variables or a variable and a sum (or subtraction) of other variables, for example:

a == b
a == (b + c)

I would have thought that taking for example these 2 vars:

a = 10.63;
b = “10.63”;

and asking

a == Number(b)

or

Number(a) == Number(b)

would have told me true, but nope! And here comes the nightmare.

I extrapolated some of the controls from my original project, and after a lot of testing i came up with this test sheet, whose results have left me impressed:


a = 10.63;
b = "10.63";

trace("a = " + a );
trace("b = \"" + b + "\"");
trace(newline);

trace("a == b ? " + (a == b));
trace("(\"\"+a) == b ? " + ((""+a) == b));
trace("a == Number(b) ? " + (a == Number(b)));
trace("typeof(a) = " + typeof(a) + "; typeof(Number(b)) = " + typeof(Number(b)));
trace(newline);
trace("Number(a) == Number(b) ? " + (Number(a) == Number(b)));
trace("Number(a) - Number(b) ? " + (Number(a) - Number(b)));
trace("typeof(Number(a)) = " + typeof(Number(a)) + "; typeof(Number(b)) = " + typeof(Number(b)));
trace(newline);
trace("a == parseFloat(b) ? " + (a == parseFloat(b)));
trace("a - parseFloat(b) ? " + (a - parseFloat(b)));
trace(newline);
trace("parseFloat(a) == parseFloat(b) ? " + (parseFloat(a) == parseFloat(b)));

trace(newline);
trace("/******************************************************************/");
trace(newline);

c = 10.63;
d = "10";
e = "0.63";

trace("c = "+c);
trace("d = \"" + d + "\"");
trace("e = \"" + e + "\"");
trace(newline);

trace("Number( d ) + Number( e ) = "+(Number( d ) + Number( e )));
trace("c == Number( d ) + Number( e ) ? "+(c == (Number( d ) + Number( e ))));
trace("Number(c) == Number( d ) + Number( e ) ? "+(Number(c) == (Number( d ) + Number( e ))));

trace(newline);
trace("/******************************************************************/");
trace(newline);

f = 10.63;
g = "10";
h = "0.63";

trace("f = "+f);
trace("g = \"" + g + "\"");
trace("h = \"" + h + "\"");
trace(newline);

trace("parseFloat(g) + parseFloat(h) = "+(parseFloat(g) + parseFloat(h)));
trace("f == parseFloat(g) + parseFloat(h) ? "+(f == (parseFloat(g) + parseFloat(h))));
trace(newline);
trace("parseFloat(f) == parseFloat(g) + parseFloat(h) ? "+(parseFloat(f) == (parseFloat(g) + parseFloat(h))));
trace("parseFloat(f) - (parseFloat(g) + parseFloat(h)) = "+(parseFloat(f) - ((parseFloat(g) + parseFloat(h)))));
trace(newline);
trace("parseFloat(f) == parseFloat(parseFloat(g) + parseFloat(h)) ? "+(parseFloat(f) == parseFloat(parseFloat(g) + parseFloat(h))));
trace("parseFloat(f) - parseFloat(parseFloat(g) + parseFloat(h)) = "+(parseFloat(f) - parseFloat(parseFloat(g) + parseFloat(h))));
trace(newline);
trace("parseFloat(f) == parseFloat(Number(g) + Number(h)) ? "+(parseFloat(f) == parseFloat(Number(g) + Number(h))));
trace("parseFloat(f) - parseFloat(Number(g) + Number(h)) = "+(parseFloat(f) - parseFloat(Number(g) + Number(h))));

These are the trace results:


a = 10.63
b = "10.63"


a == b ? false
(""+a) == b ? true
a == Number(b) ? false
typeof(a) = number; typeof(Number(b)) = number


Number(a) == Number(b) ? false
Number(a) - Number(b) ? 1.77635683940025e-15
typeof(Number(a)) = number; typeof(Number(b)) = number


a == parseFloat(b) ? false
a - parseFloat(b) ? 1.77635683940025e-15


parseFloat(a) == parseFloat(b) ? true


/******************************************************************/


c = 10.63
d = "10"
e = "0.63"


Number( d ) + Number( e ) = 10.63
c == Number( d ) + Number( e ) ? true
Number(c) == Number( d ) + Number( e ) ? true


/******************************************************************/


f = 10.63
g = "10"
h = "0.63"


parseFloat(g) + parseFloat(h) = 10.63
f == parseFloat(g) + parseFloat(h) ? true


parseFloat(f) == parseFloat(g) + parseFloat(h) ? false
parseFloat(f) - (parseFloat(g) + parseFloat(h)) = -1.77635683940025e-15


parseFloat(f) == parseFloat(parseFloat(g) + parseFloat(h)) ? true
parseFloat(f) - parseFloat(parseFloat(g) + parseFloat(h)) = 0


parseFloat(f) == parseFloat(Number(g) + Number(h)) ? true
parseFloat(f) - parseFloat(Number(g) + Number(h)) = 0

As you can see, examinating the results these are the dont’s:

  1. cannot use typeof to determine whether variables need conversion, because with 2 numbers which seem identical the equality control may fail;

  2. cannot use parseFloat on a previously declared Number variable or you will get a compilation error (function parseFloat accepts only String as an input) but the incredible thing is that if you define a Number variable without declaring its type and you parseFloat it it works.

  3. Number(a) == Number(b) fails, Number© == Number(d) + Number(e) works.

  4. parseFloat(a) == parseFloat(b) works, parseFloat(f) == parseFloat(g) + parseFloat(h) fails.

Told this, and trying to have some kind of coherence in my code, which system should i use for converting my variables either if i have to compare 2 single variables or a variable and a sum of variables?

Seems like in the first case i should use parseFloat and in the second i should use Number, which seems no sense to me.

Together with trying to write a coherent code, i’d also like to be sure that parsing an already numeric variable with parseFloat wouldn’t give me some kind of problem, even if done at runtime doesn’t give compilation errors.

Thanks in advance for any opinion you may express :wink: