C# regex formula

Question

For the following line, I need a regex to get the values outside the double quotes, namely: 0.0 and 100.5.

"VAL_ 344 PedalPos 0.0 \"% (0.0 ... 100.0)\" 100.5 \"Invalid - Undefined (100.5 ... 127.5)\";"

Using this rule Regex.Match(line, "\"\\s[0-9]+\\s\""), I am getting one group and that's the first value: 0.0. I can't figure out how to extend the search to include all the following values.

Taking into consideration this part [0-9], I think this only applies to integer values, I've added a dot there [0-9.] and this included the entire double numbers. Is this the correct way to go?

@stribizhev, no .. that value is an Id. It shouldn't be added. — Olaru Mircea, May 29 '15 at 08:05
And what is the criterion for the IDs? Always after a `VAL_`? Please clarify. — Wiktor Stribiżew, May 29 '15 at 08:06
\s?\d+(\.\d{1,2}) this should work, tested here ... http://regexpal.com/ — Dreamweaver, May 29 '15 at 08:08
@stribizhev a format like this : VAL_ 344 PedalPos is found on every line, VAL_ as a constant string, an integer value as the ID and the next is a string, meaning the friendly name for that ID. And after them, pairs like : 0.0 "% (0.0 ... 100.0)" or 100.5 "Invalid - Undefined (100.5 ... 127.5)" are found. — Olaru Mircea, May 29 '15 at 08:09
In your case the following regex will work: `([\d\.]+)\s+\\"` After character escaping: `([\\d\\.]+)\\s+\\\\"` You will get two captured groups, which you can access by index. I've checked it on http://regexr.com/ — kreig, May 29 '15 at 08:12

score 1 · Answer 1 · answered May 29 '15 at 08:07

1

Try "\s(\d+\.?\d*)\s" (string regex = "\"\\s(\\d+\\.?\\d*)\\s\""; in code) and take the first group's result.

answered May 29 '15 at 08:07

Binkan Salaryman

3,008
1
17
29

Wiktor Stribiżew · Accepted Answer · 2015-05-29T09:05:41.427

1

I suggest the following approach:

1) Remove all the quoted strings,

2) Extract all numbers that are not preceded by VAL_.

var txt = "VAL_ 344 PedalPos 0.0 \"% (0.0 ... 100.0)\" 100.5 \"Invalid - Undefined (100.5 ... 127.5)\";";
txt = Regex.Replace(txt, @"""[^""]*""", string.Empty);
var results = Regex.Matches(txt, @"(?<!VAL_\s+)-?\b\d*\.?\d+\b");

Output:

enter image description here

Regex explanation:

"[^"]*" - Match a quoted string
(?<!VAL_\s+)\b\d*\.?\d+\b:
- (?<!VAL_\s+) - A negative lookbehind to check if the number is not preceded with the constant VAL_ string and 1 or more spaces
- \b\d*\.?\d+\b - Match a whole word that is a floating number (a bit simplified, but it will even work with .04-like values).

edited May 29 '15 at 09:05

answered May 29 '15 at 08:12

Wiktor Stribiżew

607,720
39
448
563

I hope my explanation is good. And the question is by no means a duplicate. – Wiktor Stribiżew May 29 '15 at 08:17
@Rawling: Downvoting valid answers is BAD. I test my answers before posting. You do not like the question? Downvote the question. – Wiktor Stribiżew May 29 '15 at 08:22
1

I am sorry that some of the important details were mentioned in the comments. Indeed i didn't pay enough attention, i was just concentrated to my values and missed that first integer after the VAL_. But this answer gives the expected result and i thank stribizhev twice, once for its formula and also for his question about the 344 value. – Olaru Mircea May 29 '15 at 08:28
Late question: i can't find the way to get negative values also. In this case -100.5. – Olaru Mircea May 29 '15 at 09:01
1

I have added the support for negative values by adding an optional minus sign to the regex: `@"(?<!VAL_\s+)-?\b\d*\.?\d+\b"`. – Wiktor Stribiżew May 29 '15 at 09:06
i've tried that -? everywhere but in front of that boundary ..thanks again. – Olaru Mircea May 29 '15 at 09:08

score 1 · Answer 3 · answered May 29 '15 at 10:09

A more generic approach that uses a single expression to get the numbers you need as I understand it:

@"VAL_\s*\d+|""[^""]+""|(\d+(?:\.\d+)?)"

How is works is that it actually matches the parts you don't want first without doing much to it, and when it comes to the last part, it uses a capture group to get what you actually need. Here's a snippet for how to use it:

string text = "VAL_ 344 PedalPos 0.0 \"% (0.0 ... 100.0)\" 100.5 \"Invalid - Undefined (100.5 ... 127.5)\";";
var re = new Regex(@"VAL_\s*\d+|""[^""]+""|(\d+(?:\.\d+)?)", RegexOptions.IgnoreCase);
var textmatches = re.Matches(text);
Console.WriteLine("Result:");
foreach (Match match in textmatches)
{
    Console.WriteLine(match.Groups[1].Value);
}

ideone demo

VAL_\s*\d+ matches VAL_ followed by optional spaces and digits for the IDs,

""[^""]+"" gets all within double quotes,

(\d+(?:\.\d+)?) and finally this gets the numbers. I used a basic one, so if you have more complex numbers (negatives, scientific format, etc), you'll have to change that accordingly.

This ought to be more efficient than using two regex and works even if what you have between parentheses contain more than 2 numbers separated by spaces. — Jerry, May 29 '15 at 10:58

C# regex formula

3 Answers3