Extracting double value from a string c#

Question

I want to extract the double from my string.

buff = "VA VV_CELL1 3.55"

When i use the following code

private void GetLine(string msg, string buff, double numb)
{
    comPort.WriteLine(msg); 
    Thread.Sleep(50);
    buff = comPort.ReadExisting();
    Thread.Sleep(50);
    MatchCollection matches = Regex.Matches(buff, @".*?([-]{0,1} *\d+.\d+)");
    List<double> doubles = new List<double>();
    foreach (Match match in matches)
    {
        string value = match.Groups[1].Value;
        value = value.Replace(" ", "");
        doubles.Add(double.Parse(value));
        Thread.Sleep(200);
        numb = doubles[0];                
    } 
}

This code work for my other strings but "CELL1" contains a number so i dont get the wanted value "3.55" any ideas?

I don't understand the goal of Thread.Sleep in your foreach loop — Pac0
– Pac0, Commented Dec 19, 2017 at 12:24
just test the value for any nonnumerical chars(excluding the decimal seperator) before parsing — Robbert Draaisma
– Robbert Draaisma, Commented Dec 19, 2017 at 12:26
I assume you want to capture both doubles and integers, otherwise you could write \d+\.\d+. You can capture a decimal preceded by whitespace with \s+\d+(\.\d+)?. If you want to capture only the last field, \s+\d+(\.\d+)?$ — Panagiotis Kanavos
– Panagiotis Kanavos, Commented Dec 19, 2017 at 12:33

Tim Schmelter · Accepted Answer · 2017-12-19 13:06:37Z

6

Why you don't simply split this string and take the last part?

string numberPart = buff.Split().Last();
double num;
bool validNum = double.TryParse(numberPart, NumberStyles.Any, CultureInfo.InvariantCulture, out num);

Another way is to use Substring and LastIndexOf(which fails if there is no space):

string numberPart = buff.Substring(buff.LastIndexOf(' ')).Trim();

To help on your comment:

I'd use a method that returns a double?(double that can be null):

double? GetNumber(string buff)
{
    string numberPart = buff.Split().Last();
    double num;
    bool validNum = double.TryParse(numberPart, NumberStyles.Any, CultureInfo.InvariantCulture, out num);
    if (validNum)
        return num;
    else
        return null;
}

Now you can use the method and you even know whether the number could be parsed successfully or not:

double? result = GetNumber("VA VV_CELL1");
bool wasValid = result.HasValue;
if(wasValid)
{
   double value = result.Value;
}

edited Dec 19, 2017 at 13:06

answered Dec 19, 2017 at 12:22

Tim Schmelter

462k79 gold badges718 silver badges980 bronze badges

Sign up to request clarification or add additional context in comments.

8 Comments

Panagiotis Kanavos Over a year ago

Splitting generates a lot of temporary strings that can be very expensive in when parsing a large number of records. If the record is fixed length though, the lines short or the number of records small, regular expressions can be overkill

Tim Schmelter Over a year ago

@PanagiotisKanavos: In OP's sample it generates 3 temp strings. I dont think this is too much overhead. However, you can also use buff.Substring(buff.LastIndexOf(' ')).Trim(). Normally regex is much more expensive

xBehemoth Over a year ago

Thx mate it helped a lot!

xBehemoth Over a year ago

@TimSchmelter ` private void GetGet(string msg, string buff, double var) { comPort.WriteLine(msg); Thread.Sleep(200); buff = comPort.ReadLine(); Thread.Sleep(200); string numberPart = buff.Split().Last(); double num; bool validNum = double.TryParse(numberPart, NumberStyles.Any, CultureInfo.InvariantCulture, out num); var = num; } ` when i call my method GetGet("VA VV_CELL1", buff1, var1); var1 is 0, why?

Tim Schmelter Over a year ago

@xBehemoth: you either have to return num or make the parameter var ref. I'd prefer a method double GetValue(string msg, string buff, double var){ ... return num;}

|

Marco Antonio · Accepted Answer · 2017-12-19 12:33:29Z

1

Try, this regex expression : \s+\d+(.)?\d+

answered Dec 19, 2017 at 12:33

Marco Antonio

211 bronze badge

Comments

Panagiotis Kanavos · Accepted Answer · 2017-12-19 12:50:58Z

I assume you want to capture both doubles and integers, otherwise you could write \d+\.\d+. This :

Regex.Matches("VA VV_CELL1 3.55",@"\d+\.\d+")[0]

Returns 3.55 but can't capture 355.

You can capture an integer or decimal preceded by whitespace with \s+\d+(\.\d+)?.

Regex.Matches("VA VV_CELL1 3.55",@"\s+\d+(\.\d+)?")[0]

Returns 3.55 while

Regex.Matches("VA VV_CELL1 355",@"\s+\d+(\.\d+)?")[0]

Returns 355

If you want to capture only the last field you can use \s+\d+(\.\d+)?$,eg:

Regex.Matches("VA VV_CELL1 3.54 3.55",@"\s+\d+(\.\d+)?$")[0]

Returns 3.55

You don't need to trim whitespace because double.Parse ignores it. You can change the pattern to capture the number in a separate group though, by surrounding the digits with parentheses :

Regex.Matches("VA VV_CELL1 3.54 3.55",@"\s+(\d+(\.\d+)?)$")[0].Groups[1]

You need to use Groups[1] because the first group always returns the entire capture

Collectives™ on Stack Overflow

Extracting double value from a string c#

3 Answers 3

8 Comments

Comments

Comments

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

8 Comments

Comments

Comments

Related