C# Linq 从结果中删除字符以进行比较

Question

I have a linq query which fetches records from SQL server database. Phone numbers for example are stored as strings and can contain spaces or characters that when searching are not found.

ie I search for a phone number "0123456789" but in the database it's stored as "01234 567890" or "01234-567890" etc. In my code below you can see my attempts to strip out the characters (right or wrongly!) but I still do not get any results unless I search for the characters after the empty space "567890".

With the code below:

"567890" returns correct results

"01202 567890" returns no results

"01202567890" returns no results

    Expression<Func<Customer, bool>> SearchVals = ContainsInDescription("01234567890");

    var query = (from customer in data.Customers.Where(SearchVals)
                                 where customer.Deleted == false
                                 orderby customer.Name ascending
                                 select new
                                 {
                                     CustomerID = customer.CustomerID.ToString(),
                                     Name = customer.Name,
                                     Company = customer.Company,
                                     Home = customer.PhoneHome.Replace(" ", ""),
                                     Mobile = customer.PhoneMobile.Replace(" ", ""),
                                     Work = customer.PhoneWork.Replace(" ", ""),
                                     Email = customer.Email,
                                     Address1 = customer.Address.AddressLine1,
                                     Address2 = customer.Address.AddressLine2,
                                     City = customer.Address.City,
                                     County = customer.Address.County,
                                     Postcode = customer.Address.PostalCode,
                                     Country = customer.Address.Country
                                 }).Skip(totalToDisplay * page).Take(totalToDisplay);

public Expression<Func<Customer, bool>> ContainsInDescription(
                                                        params string[] keywords)
        {
            Expression<Func<Customer, bool>> predicate = PredicateBuilder.False<Customer>();
            foreach (string keyword in keywords)
            {
                string temp = keyword;

                //string predicate1 = PhoneHome.ToString().Replace(" ", "").Replace("-", "").Replace("(", "").Replace(")", "").Trim();
                string predicate2 = temp.ToString().Replace(" ", "").Replace("-", "").Replace("(", "").Replace(")", "").Trim();

                predicate = predicate.Or(p => p.Name.Contains(temp));
                predicate = predicate.Or(p => p.Company.Contains(temp));

                predicate = predicate.Or(p => p.PhoneHome.Contains(temp));
                predicate = predicate.Or(p => p.PhoneMobile.Contains(temp));
                predicate = predicate.Or(p => p.PhoneWork.Contains(temp));
                //predicate = predicate.Or(p => p.PhoneHome.ToString().Replace(" ", "").Replace("-", "").Contains(temp.ToString().Replace(" ", "").Replace("-", "")));
                //predicate = predicate.Or(p => p.PhoneMobile.ToString().Replace(" ", "").Replace("-", "").Contains(temp.ToString().Replace(" ", "").Replace("-", "")));
                //predicate = predicate.Or(p => p.PhoneWork.ToString().Replace(" ", "").Replace("-", "").Contains(temp.ToString().Replace(" ", "").Replace("-", "")));
                //predicate = predicate.Or(p => p.PhoneHome.ToString().Replace(" ", "").Replace("-", "").Replace("(", "").Replace(")", "").Trim() == temp.ToString().Replace(" ", "").Replace("-", "").Replace("(", "").Replace(")", "").Trim());
                //predicate = predicate.Or(p => p.PhoneMobile.ToString().Replace(" ", "").Replace("-", "").Replace("(", "").Replace(")", "").Trim() == temp.ToString().Replace(" ", "").Replace("-", "").Replace("(", "").Replace(")", "").Trim());
                //predicate = predicate.Or(p => p.PhoneWork.ToString().Replace(" ", "").Replace("-", "").Replace("(", "").Replace(")", "").Trim() == temp.ToString().Replace(" ", "").Replace("-", "").Replace("(", "").Replace(")", "").Trim());

                predicate = predicate.Or(p => p.Email.Contains(temp));
                predicate = predicate.Or(p => p.Address.AddressLine1.Contains(temp));
                predicate = predicate.Or(p => p.Address.AddressLine2.Contains(temp));
                predicate = predicate.Or(p => p.Address.City.Contains(temp));
                predicate = predicate.Or(p => p.Address.County.Contains(temp));
                predicate = predicate.Or(p => p.Address.PostalCode.Contains(temp));
                predicate = predicate.Or(p => p.Address.Country.Contains(temp));
            }
            return predicate;
        }

Answer 1

You might need to clean the data you gather from the database as well. Currently you only remove spaces. Instead of replacing lots of stuff, consider using a whitelist approach:

using System;
using System.Linq;
using System.Text;

public static void Main()
{
    Func<string, String> onlyNumbers = data 
      => data.Where(c => char.IsDigit(c))  // only keep digits
             .Aggregate(new StringBuilder(),(builder, c) => builder.Append(c))
             .ToString(); // gimme concatted string
    
    Console.WriteLine(onlyNumbers( "sadf(234)-23492sas" ));  // =>   "23423492"
}

using above "cleaner" something like

...
Mobile = onlyNumbers(customer.PhoneMobile),
Home = onlyNumbers(customer.PhoneHome),
Mobile = onlyNumbers(customer.PhoneMobile),
Work = onlyNumbers(customer.PhoneWork),
...

and

pred = pred.Or(p => p.Mobile.Contains(onlyNumbers(_some_input_)));

should do the trick.

Instead of char.IsDigit() you could define your own hashset of allowed characters (or allow '+' as well) to allow fe international pre-dial numbers given as +41 1234 5678 90

Caveat - this will generate numebers from any text and might give you false-positives:

onlyNumbers("This 246 cars had 102 violations of rule 47.11 in the year 2012");

gives you a string of "24610247112012" .