Need a regex to extract all parameters from string.Format text

I want to create a SafeStringFormat which makes sure the parameter count matches the embedded parameters in the string text for a String.Format.

I'm guessing regex is the most efficient way to go; could somebody give me the code for extracting the parameters?

Something like this:

public string SafeStringFormat(string text, params object[] parms) {

    string[] embeddedParameters = <use regex to extract parameters from text>
    foreach (object o in parms) {
         if (<o doesn't corresponds at least one of the parameters in embeddedParameters>) {
             <throw exception saying that parm[#] is not implemented in <text>>
         }
     }
     foreach (string s in embeddedParameters) {
         if (<s doesn't correspond to one of the objects in parms>) {
             <throw exception saying s is an invalid parameter>
         }
     }
     return string.Format(text, parms);
}

It's the
string[] embeddedParameters = <use regex to extract parameters from text>
part of the code that I need help with...
LVL 1
FrancineTaylorAsked:
Who is Participating?
 
Todd GerbertIT ConsultantCommented:
Regex's are tricky.  This is closer, but still not quite right:
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Text.RegularExpressions;

class Program
{
	static void Main(string[] args)
	{
		string tests = @"This has four parms {0}{1}{2}{3}
out of sequence types {42}{99}
...but these were not extracted correctly:
{0:D}
{0,10:C}
Hello {0}, {{ , {1:D}, }}, {{3}, {4}}";

		foreach (string s in tests.Split(new string[] { "\r\n" }, StringSplitOptions.None))
			ShowParameterInfo(s);

		Console.Write("Press any key to exit...");
		Console.ReadKey();
	}

	static void ShowParameterInfo(string formatString)
	{
		Regex regex = new Regex(@"[.-[\{]]*\{(\d+)([^}]*)}[.-[}]]*");

		MatchCollection matches = regex.Matches(formatString);

		Console.WriteLine("Input string: {0}", formatString);
		Console.WriteLine("Parameter count: {0}", matches.Count);

		foreach (Match m in matches)
		{
			Console.WriteLine("\tParameter number: {0}, additional parameter details: {1}",
				m.Groups[1].Value, m.Groups[2].Value);
		}

		Console.WriteLine();
	}
}

Open in new window


Using String.Format("{{0}", "Hello") is invalid, but String.Format("{{{0}", "Hello") and would produce {Hello.  I'm not quite sure how to ensure that a { doesn't immediately precede {0}, unless {{ precede's the {0}, using regular expressions...
0
 
jdavistxCommented:
I would think a regex as simple as
(\{\d+\})

Open in new window

would give you your matches, but you'd have to also validate that your numbers are parameter numbers are sequential, and start at 0.

Not to be unhelpful, but I believe you'll get a compiler error if you try to use the normal String.Format() with invalid parameters.  You aren't really making it "safe" with what you've set out initially.

Also, I would think that you wouldn't want to throw exceptions in this way.  I've always considered Exceptions as things that are... exceptions, and not situations that can't be appropriately handled.  An exception would be some program-catastrophic event that no code would be able to accommodate.  In your sample, you could perform preemptive tests that would circumvent those "exceptions" from ever occurring.
0
 
jdavistxCommented:
If only I could edit!

I would think a regex as simple as

1:

      

(\{\d+\})

Toggle HighlightingOpen in New WindowSelect All

would give you your matches, but you'd have to also validate that your parameter numbers are sequential, and start at 0.

Not to be unhelpful, but I believe you'll get a compiler error if you try to use the normal String.Format() with invalid parameters.  You aren't really making it "safe" with what you've set out initially.

Also, I would think that you wouldn't want to throw exceptions in this way.  I've always considered Exceptions as things that are... exceptions, and not situations that can be appropriately handled.  An exception would be some program-catastrophic event that no code would be able to accommodate.  In your sample, you could perform preemptive tests that would circumvent those "exceptions" from ever occurring.
0
Get expert help—faster!

Need expert help—fast? Use the Help Bell for personalized assistance getting answers to your important questions.

 
Todd GerbertIT ConsultantCommented:
I think "[^\{]\{(\d+)[^}]*}[^}]" will work better - as you need to account for two consecutive curly-braces being treated as an escape, meaning {{0} or {0}} should not be treated as a parameter position.

I also know String.Format will throw an exception if there aren't enough objects to fill all the {0} placeholders, but the reverse of that is not true, e.g. String.Format("{0}", "one", "two", "three") will not raise exception.

I think it is also appropriate to raise an exception here, since this method should handle just it's assigned task and not need to worry about how or if to correct the error.  The code calling this method should wrap it in a try/catch block and handle errors there if necessary.

The numbers in the {0} placeholders don't need to be sequential, although the parameter array must have at least N+1 elements, where N is the highest numbered placeholder.

Regex regex = new Regex(@"[^\{]\{(\d+)[^}]*}[^}]");

string test = "Hello {0}, {{ , {1:D}, }}, {{3}, {4}}";
MatchCollection matches = regex.Matches(test);

Open in new window

0
 
Todd GerbertIT ConsultantCommented:
On exceptions...here the calling code handles the exception - you wouldn't want for SafeStringFormat to be showing message boxes and such.  SafeStringFormat might end up in a class library, in which case you have no idea how exceptions should be handled in future circumstances.

Alternatively, you could have the SafeStringFormat return a bool or int instead of raising an exception - though I think raising the exception would be more in line with the preferred .Net model.

Also maybe consider one method that raises exceptions, and another that returns a bool - a la Int32.Parse() and Int32.tryParse()

static void Main()
{
retry:
  try
  {
    SafeStringFormat("Hello {0}", "one", "two", "three");
  }
  catch (InvalidFormatException ex)
  {
    if (MessageBox.Show("Invalid format, try again?") == DialogResult.Yes)
      goto retry;
  }
}

static void SafeStringFormat(string format, params object[] args)
{
  if (args.Length > NumberOfPlaceHolders)
    throw new InvalidFormatException("Not enough placeholders.");
  else if ( args.Length < NumberOfPlaceHolders)
    throw new InvalidFormatException("Too many placeholders.");
}

Open in new window

0
 
FrancineTaylorAuthor Commented:
Okay, here is the code I used to test the regex:

            MatchCollection matches = regex.Matches(text);
            List<string> list = new List<string>();
            foreach (Match match in matches) {
                list.Add(match.Groups[1].Value);
            }
            string[] embeddedParameters = list.ToArray();

When I used it with jdavistx's regex ("(\{\d+\})"), I got the correct result from these strings:
    "This has four parms {0}{1}{2}{3}"
    "out of sequence types {42}{99}"
...but these were not extracted correctly:
    "{0:D}"
    "{0,10:C}"
    "Hello {0}, {{ , {1:D}, }}, {{3}, {4}}"

tgerbert's regex ("[^\{]\{(\d+)[^}]*}[^}]") didn't appear to extract any of them correctly.

Is that an accurate test, or am I using the MatchCollection incorrectly?
0
 
FrancineTaylorAuthor Commented:
That works perfectly!  Thanks, tgerbert...
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.