Parsing a text file using C#

Hi  Gerry,

Thanks yes that was I have done too. However, I had to do some cleaning since for situations like:

 Date: 6/7/2009                        Test Form: A23

I had to check the position of "Date:" and then extracting after ":" or same thing for "Test From:" etc using  IndexOf() and Substring() methods.

Any idea if I'm on a right track?

Thanks.

That will do nicely and should work fine.
 
...or you could get clever with .Split() and LINQ   :)

        static void Main(string[] args)
        {
            string text = " Date: 6/7/2009                        Test Form: A23";
            System.Console.WriteLine("Original text: '{0}'", text);
 
            var words = from w in text.Split(' ',':','\t').AsEnumerable<string>() where w!=string.Empty select w;
            
            System.Console.WriteLine("{0} words in text:", words.Count());
 
            foreach (string s in words)
            {
                System.Console.WriteLine(s);
            }
            Console.ReadLine();
        }
                                              

1:
2:
3:
4:
5:
6:
7:
8:
9:
10:
11:
12:
13:
14:
15:

Thanks for your comment. I don't know anything about Linq since I'm new to C# but will check it out for sure.

I will get back to you.

Thanks!

For linq you need C# 3.5

another way is to use regular expressions (Regex, System.Text.RegularExpressions).

What's important is that you define the syntax rules. Here's a sequential set of rules that may make sense:

Any character except ":" = first field name
":"
Any character except tab (\t) = first value
"\t"
Any character except ":" = second field name
":"
All other characters

This could be defined in regex like the following. It also strips spaces and names the capture groups  1=name1, 2=value1, 3=name2, 4=value2

 *(?<name1>[^:]*) *: *(?<value1>[^\t]*) *\t *(?<name2>[^:]*) *: *(?<value2>.*) *

                                              

1:

Hello,

Thanks for your comments. I have attached the format I am receiving (after converting a specific file to text). Is the above method still good for it or should I change my approach?

In following example I will need to extract;
William Smith
5/2/2008
55 ( which is exam score)
1025409804

Once again thanks.

Regards,

Header file
Name of visitor: William Smith Signin Date: 5/2/2008 Position Applied For: Driver Number Correct (exam score): 55 Percentile Total (%total): 78 Median Score for Position: 52 Applicant ID: 1025409804 Test Form: ZipCode 1x20558 Job Code: 000001 Age Adjusted Score: 52 Equiv: 117 Suggested Hiring Range: 19 - 44

                                              

1:
2:
3:

I just found out I'm using (base on help dialog) .NET 3.5 sp1

so I guess I can use Linq right?

Yup, or regular expressions, or plain old IndexOf/Substring... that's why programming is so fun.
Your attached example is a bit tricky.... trying to separate the "William Smith" from "Signin" etc.  
You need to know in the code the full label names, making it very tricky for an otherwise excellent regular expression approach but easy for IndexOf/Substring and perhaps the rather terse Linq example I posted earlier.

In the interest of the principle of KISS (except I'm sure you're not stupid :-) maybe IndexOf/Substring is the best approach after all !

Will the field headings always be the same?

Name of visitor
Signin Date
Position Applied For
Number Correct (exam score)
Percentile Total (%total)
Median Score for Position
Applicant ID
Test Form
Job Code
Age Adjusted Score
Equiv
Suggested Hiring Range

This would make life easier.

I don't have time now. I'll try and put a Regex script together later.

Yes, it will be always like that.

Thanks.

I've just noticed your latest example is different to the original example.

Will all the fields ALWAYS be present and in the same order?
Will a single entry cover multiple lines?
Are you talking about one entry per file?
is "Header file" part of the data to parse?

Hi Tiggerito,

Yes, consider the last one since I'm generating them as the latter one.