Hi

I’m using Regex to split values from csv file (using C#) .

Code on this page: http://regexadvice.com/blogs/wayneking/ do the job.

Generally it splits values for some delimiter (in my case  comma {,}) and also correctly split the values that are surrounded by double quotes (case when comma is part of content not as delimiter).

The split is performed by following pattern:

@"""([^""\\]*[\\.[^""\\]*]*)""" +

         "|" +

         @"([^" + delimiters + @"]+)"

Where ‘delimiters’ is parameter passed to function and define delimiters.

I’m using just one {,}.

This is complete function:

public static string[] SplitQuoted(string text, string delimiters)

    {

        // Default delimiters are a space and tab (e.g. " \t").

        // All delimiters not inside quote pair are ignored. 

        // Default quotes pair is two double quotes ( e.g. '""' ).

        if (text == null)

            throw new ArgumentNullException("text", "text is null.");

        if (delimiters == null || delimiters.Length < 1)

            delimiters = " \t"; // Default is a space and tab.

        ArrayList res = new ArrayList();

        // Build the pattern that searches for both quoted and unquoted elements

        // notice that the quoted element is defined by group #2 (g1)

        // and the unquoted element is defined by group #3 (g2).

        string pattern =

         @"""([^""\\]*[\\.[^""\\]*]*)""" +

         "|" +

         @"([^" + delimiters + @"]+)";

        // Search the string.

        foreach (System.Text.RegularExpressions.Match m in System.Text.RegularExpressions.Regex.Matches(text, pattern))

        {

            string g0 = m.Groups[0].Value;

            string g1 = m.Groups[1].Value;

            string g2 = m.Groups[2].Value;

            //if (g2 != null && g2.Length > 0)

            //{

            //    res.Add(g2);

            //}

            //else

            //{

            //    // get the quoted string, but without the quotes in g1;

            //    res.Add(g1);

            //}

            res.Add(g2);

        }

        return (string[])res.ToArray(typeof(string));

    }

Everything works just fine when all values in all rows are not empty, but when some value is empty it will not match it.

So for example:

One,two,three will return 3 matches

But

One,,three will return 2 matches

How to adopt this regex pattern so It will match empty string. I’m not that good with regular expressions but believe that this group @"([^" + delimiters + @"]+)" in pattern is responsible for that .

Any help is appreciated