ex

A lightweight, ruby-jealous c# extension class that makes regular expressions fun again.

What It Does

1. strings have been suped up with Regex-powered methods HasPattern, MatchesPattern, Sub, Scan, (and more!)
2. The MatchesPattern method returns a new class MatchData that simplifies access to captures. It's a little like ruby's class of the same name. See this answer on stackoverflow for a better explanation of why the default implementation of MatchCollection is so complicated.
3. Any RegexOptions can be expressed as a character string
4. When using .NET 4, regex objects are cached

What It Does Not Do

1. This is not a System.Text.RegularExpressions replacement. It just greases the wheels a little.
2. Compile on framework versions below 3.5

Examples

Test a match

"Adam & Steve".HasPattern(
    @"Adam (&|and) (?<someone_else>\w+)"
); // Returns true

Test a match with case insensitivity option:

"ADAM AND STEVE".HasPattern(
    @"adam (&|and) (?<someone_else>\w+)",
    "i"   // RegexOptions are expressed as a character string (see below for reference)
); // Returns true

Find the first match

"haystack needle haystack".FindPattern(@"\s(needle)\s");  // returns " needle "

Find the first match, but return the first capture group.

"haystack needle haystack".FindPattern(@"\s(needle)\s", 1);  // returns "needle"

Substitution:

// Replace all with GSub
"A man, a plan, a canal, panama".GSub("a.", "ax");  // "A max, axplax, axcaxax, paxaxa"

// Replace first with Sub
"A man, a plan, a canal, panama".Sub("a.", "ax");  // "A max, a plan, a canal, panama"

// Replace all using a lambda evaluator (no need for MatchEvaluator delegate!)
var actors = "Alec Baldwin, Daniel Baldwin, William Baldwin";
actors.GSub(
    @"(?<firstname>\w+)\s(?<lastname>\w+)",
    match =>
    {
        // Replaces any name found with "Steven Baldwin"
        return "Steven Baldwin";
    }
); // returns "Steven Baldwin, Steven Baldwin, Steven Baldwin"

Scan:

Scan calls the supplied function for each pattern match.

string sentence = @"In west Philadelphia born and raised
    On the playground where I spent most of my days"

sentence.Scan(@"\w+", w =>
{
    // called once for each word, with 'w' containing the match value.
});

Captures and capture groups, if used, are given as parameters!

string response = @"HTTP/1.1 200 OK
    Server: nginx/1.0.4
    Date: Fri, 24 Jun 2011 21:52:36 GMT
    Content-Type: text/html; charset=utf-8
    Transfer-Encoding: chunked
    Connection: keep-alive
    Status: 200 OK
    Etag: 924990f60843c36a22f65ec789ea33f3
    X-Runtime: 8ms
    Cache-Control: private, max-age=0, must-revalidate
    Strict-Transport-Security: max-age=2592000
    Content-Encoding: gzip"

response.Scan(@"([a-zA-Z\-]+): (.+)", (name, value) =>
{
    // name, value are both strings containing the next captured header
});

Match Data

MatchData is a replacement for MatchCollection except it's actually a joy to use. You can use it as a dictionary of captures or to retrieve capture details. It only has a minor limitation compared to MatchCollection, and that limitation is a corner case.

string fullname = "Lee Harvey Oswald";
var m = fullname.MatchesPattern(@"(?<firstname>\w+)\s(\w+)\s(?<lastname>\w+)");

// m["firstname"] == "Lee"
// m["lastname"] == "Oswald"
// m[0] == "Lee Harvey Oswald"  (The first element is always the entire match [ask Larry Wall])
// m[1] == "Harvey"             (The second element is always the first numbered match)

// m.Begin(1) == 4              (This is the beginning offset of "Harvey")
// m.End(1) == 10               (This is the end offset of "Harvey")

// m.Begin("lastname") == 11    (This is the beginning offset of "Oswald")
// m.End("lastname") == 17      (This is the end offset of "Oswald")

Isn't this refreshing?

Option Reference

i: Ignore case
s: Singleline mode (period [.] matches newlines)
x: Ignore pattern whitespace
c: compile pattern
r: right to left evaluation

Thank you to SLaks at codereview.stackexchange.com for his helpful review.