How to convert Wildcard pattern into regex pattern using C#

Question

For eg:

I have wildcard pattern to find UK format dates

Wildcard pattern : [0-9]{1,2}[dhnrst]{2} <[AFJMNSOD]*> [0-9]{4} // this pattern will find the UK format date "7th August 2001"

Out put also will find the UK format dates, but out put should be in regex format as follows.

\b[0-9]{1,2}[th|rd]+\s[\w]+\s[0-9]{4}\b // this regex pattern also will find the UK date format "7th August 2001"

Please suggest me how to bring this output using c#. (Except string find and replacement method)

Vulpes · Answer

The following is the closest I can get to automatically mapping an MS Word wildcard expression to a regular expression: using System; using System.Collections.Generic; using System.Text.RegularExpressions; class Program { static void Main() { var mappings = new Dictionary (); mappings.Add(@"\?", "\xff"); mappings.Add("?", "."); mappings.Add("\xff", @"\?"); mappings.Add(@"\*", "\xff{1}"); mappings.Add("*", ".*"); mappings.Add("\xff{1}", @"\*"); mappings.Add(" mappings.Add(">", @"\b"); mappings.Add("[!", "[^"); mappings.Add("@", "+"); string wildcard = "[0-9]{1,2}[dhnrst]{2} [0-9]{4}"; string regex = WildcardToRegex(wildcard, mappings); Console.WriteLine(regex); bool isValid = Regex.IsMatch("7th August 2001", regex); Console.WriteLine(isValid); Console.ReadKey(); } static string WildcardToRegex(string wildcard, Dictionary mappings) { foreach(string key in mappings.Keys) { wildcard = wildcard.Replace(key, mappings[key]); } return wildcard; } } The output is: [0-9]{1,2}[dhnrst]{2} \b[AFJMNSOD].*\b [0-9]{4} True A similar approach could be used to map a regular expression to an MS wildcard expression provided the former doesn't include any features which can't be represented in the latter such as alternation, optional characters, look-ahead, look-behind etc.

Penikton Bernard · Answer

Thank you so much. I got answer dude. Thanks for your help.

Penikton Bernard · Answer

I don't have documentation yet And I am waiting to add this functions in my socurce

Penikton Bernard · Answer

The below functions are I need to add in my source Create function for: (1) Convert MS Word wildcard search string to regular expression string (2) Convert regular expression string to MS Word wildcard search string

Vulpes · Answer

That may not be possible as the wildcard format you're using is unlikely to be powerful enough to represent a complex regular expression. Do you have a link to documentation on the wildcard pattern or is the above example all you have to work from?

Penikton Bernard · Answer

I don't really want the UK format string output. I just shown as an example for you. But I really want the output shout be wildcard pattern. if I pass the parameter as string format that contains "regex pattern" then I should get the output as in the "wildcard pattern" for the regex pattern. Hope you understand my need and I hope I ill get answer form you.

Vulpes · Answer

I've just used the temporary strings s1, s2 and s3 to split up what would otherwise be an unwieldy regular expression into smaller, more understandable, parts. You can write the whole lot as a single string if you like. The reason why I say this is more comprehensive than the wildcard pattern is that the latter would validate something like: 0dn Application 0000 as a valid UK format date. So you really need something better than this.

Penikton Bernard · Answer

Hi, Am sorry for making you disturb. I referred your code but what I exactly need is the out put of result should be in "Regex" Or "Wildcard pattern" For Eg. Your are giving the samples as s1,s2,s3 that are contain regex pattern. This will match the UK date format. I agree with this. But my requirement is If I provide regex pattern to find UK words then I should get my output in Wildcard pattern to find UK format words. But your output seems to be find only the matching string. I want output as wildcard pattern. If my input regex pattern to find UK format date as follows s1= string s1 = @"((1|21|31)st|(2|22)nd|(3|23)rd|(10|11|12|13|20|30)th|[1-2]?[4-9]th)\s+(January|March|May|July|August|October|December)"; Then I should get the output in wildcard string format as follows result="[0-9]{1,2}[dhnrst]{2} [0-9]{4}"; Please suggest me to proceed further.

Vulpes · Answer

Here's a much more comprehensive regular expression for this particular date format. The only problem it has is that it will always validate February 29th even if it's not a leap year. If you don't want matches to be case insensitive, then just remove the (?i) from the regular expression. string s1 = @"((1|21|31)st|(2|22)nd|(3|23)rd|(10|11|12|13|20|30)th|[1-2]?[4-9]th)\s+(January|March|May|July|August|October|December)"; string s2 = @"((1|21)st|(2|22)nd|(3|23)rd|(10|11|12|13|20|30)th|[1-2]?[4-9]th)\s+(April|June|September|November)"; string s3 = @"((1|21)st|(2|22)nd|(3|23)rd|(10|11|12|13|20)th|[1-2]?[4-9]th)\s+(February)"; string re = "(?i)^(" + s1 + "|" + s2 + "|" + s3 + @")(\s+|,\s*)[1-9][0-9]{0,}$" ; string[] dates = {"7th August 2001", "29th February 2003", "1nd May 2006", "23rd JULY 2011", "30th february 2013", "31st june 20114", "19TH oCtober 2", "2nd October 15", "15th April, 979", "27th MAY 12113"}; foreach(string date in dates) { bool b = Regex.IsMatch(date, re); Console.WriteLine("{0} -> {1}", date, b); } The output is: 7th August 2001 -> True 29th February 2003 -> True 1nd May 2006 -> False 23rd JULY 2011 -> True 30th february 2013 -> False 31st june 20114 -> False 19TH oCtober 2 -> True 2nd October 15 -> True 15th April, 979 -> True

Penikton Bernard · Answer

Hi, Thanks a lot for looking my post, The input string should be a pattern its either regex or wildcard Its should not be the normal text

Khan Abrar Ahmed · Answer

Hi you can Refer this links http://stackoverflow.com/questions/5899794/match-and-replace http://stackoverflow.com/questions/6005609/replace-only-some-groups-with-regex http://stackoverflow.com/questions/474335/using-regex-to-edit-a-string-in-c-sharp

Vulpes · Answer

If you're looking for an exact translation of that wildcard pattern to a regular expression, then this should do it: string text = "7th August 2001"; string re = @"^[0-9]{1,2}[dhnrst]{2} [AFJMNSOD]\w+ [0-9]{4}$"; bool b = Regex.IsMatch(text, re); Console.WriteLine(b); // True However, as I'm sure you know, these patterns are far from perfect and it would be possible to come up with a more accurate regular expression if that were needed.

How to convert Wildcard pattern into regex pattern using C#

Answers (12)