正则表达式从列表中查找完全匹配

Question

I had a quick question that I was hoping someone would be able to shed some light on for me.

Still new to RegEx so this behavior doesn't make sense to me. I am using C# to write a simple function that searches for a list of substrings within a given string and that substrings position in the string. My code looks like this:

DataTable matchtable = new DataTable();
string searchstring = " Take a left in 2.1 miles.  Then take a right in 3 miles";
var substringlist = new [] {"2.1 miles", "3 miles", "4.1 miles", "1","take"};
string searchregexstr = string.(@"(\W|^){0}(\W|$)", string.Join("|", substringlist));
Regex searchregex = new Regex(searchregexstr);
if (searchregex.IsMatch(searchstring))
{
    foreach (Match substring in searchregex.Matches(searchstring))
    {
        string substringmatch = substring.toString();
        int indexofsubstringmatch = searchstring.IndexOf(substringmatch);
        matchtable.Rows.Add(susbtringmatch, indexofsubstringmatch);
    }
    return matchtable;
}
return matchtable;

With my main regex match function looking like this:

string searchregexstr = string.(@"(\W|^){0}(\W|$)", string.Join("|", substringlist));

My issue is:

When looking at my match table results I get a hit for both 2.1 miles and 1 (which is being matched within the 2.1)

I assume (I think incorrectly) that my regex is looking for only complete matches where 1 should not match because it is not found by itself in the string.

Does something stand out as missing?

Thanks very much for any and all help in advance!

Zinga

Answer 1

Well, you can do this in many ways. eg following code will return you list of indices of terms found in a particular string.

public static IEnumerable<int> GetStringIndices(IEnumerable<string> substringlist, string data)
{
    var lstIndices = new List<int>();

    foreach (var searchString in substringlist)
    {
        var regexObj = new Regex($@"(?<=(\s|^)){searchString}(?=(\s|$)|(\W)+?)", 
            RegexOptions.IgnoreCase | RegexOptions.Multiline);

        var matchResults = regexObj.Match(data);

        if (!matchResults.Success)
        {
            lstIndices.Add(-1);
            continue;
        }

        while (matchResults.Success)
        {
            var idx = matchResults.Index;
            lstIndices.Add(idx);

            matchResults = matchResults.NextMatch();
        }
    }
    return lstIndices;
}

if I pass search string and terms you mentioned above

var data = "Take a left in 2.1 miles.  Then take a right in 3 miles";
var substringlist = new[] { "2.1 miles", "3 miles", "4.1 miles", "1", "take" };
var indices = GetStringIndices(substringlist, data);

you'll get list of indices in variable named indices. end result will be

[15, 48, -1, 17, 0, 32]

2.1 miles is found at index 15

3.1 is at index 48 and so on.

Answer 2

Your code has some errors, for example, string.( , toString , susbtringmatch , and a runtime error when adding a row to a DataTable with no columns. BTW, do you really need a DataTable ?

Having corrected the typos and removed the DataTable , your code works fine for me if you correct this line like this:

string searchregexstr = string.Format(@"(\W|^){0}(\W|$)", string.Join("|", substringlist));

The matches are:

 2.1 miles (with leading space)
take  (with trailing space)
3 miles

Finally, you don't need the first return , as the final one will suffice.

If you need help tuning your regular expression, I highly recommend RegExr .