Trying to use System.Double.Parse(string) method for strings such as "-1.#IND" and "INF" representing special values results in a FormatException.
Is there any built-in .NET framework support to parse these?
Asked
Active
Viewed 371 times
5
etkarayel
- 373
- 1
- 2
- 9
-
What value you expect when you pass string -1.#IND to double.parse? I – DevelopmentIsMyPassion Feb 21 '14 at 13:51
-
@AshReva, I am quite sure [`double.NegativeInfinity`](http://msdn.microsoft.com/en-us/library/system.double.negativeinfinity.aspx). – Sinatr Feb 21 '14 at 13:52
-
Actually `double.NaN`. "-1.#IND" seems to be standard representation created by some technologies. – etkarayel Feb 21 '14 at 13:56
-
4*Some technologies* are not necessary standard and supported by .Net Framework. – Sinatr Feb 21 '14 at 13:58
-
For reference, [this question](http://stackoverflow.com/q/347920/87399) talks about these string values. It looks like they're from the C/C++ standard library. – Joe White Feb 21 '14 at 14:22
2 Answers
6
No, the only non-numeric values double.Parse recognizes are the string values returned by double.Nan.ToString(), double.PositiveInfinity.ToString(), and double.NegativeInfinity.ToString() (dependent on Culture).
In your case I would just use a switch:
double dblValue;
switch strValue
{
case "-1.#IND":
dblValue = double.Nan;
break;
case "INF":
dblValue = double.Infinity;
break;
//... other casess
default:
dblValue = double.Parse(strValue);
break;
}
D Stanley
- 149,601
- 11
- 178
- 240
2
NaN and other values are parsed in the specified culture (or neutral, if no culture is specified). You can play with those here if you want.
If you have to parse something more special, then just
public double MyParse(string text)
{
if(text == "blablabla")
return double.NaN;
if(text.Contains("blablabla")) ...
if(text.StartsWith(...
return double.Parse(text);
}
Sinatr
- 20,892
- 15
- 90
- 319