Regular expression C#

Question

Having trouble determining if a file name conforms to a specific convention as follows. Using regular expression in C# .Net 4.0.

Valid Format: xxxxT_SSS_sss[i]_t#y.png
where
// x = Any single character. // T = Digit: 1 to 7 inclusive. // _SSS = Positive Integer: 000 to 999 inclusive. Always padded with leading zeros. // _sss = Positive Integer: 000 to 999 inclusive. Always padded with leading zeros. // i = Random text of any length including any characters. Will always be enclosed in square [] brackets. Optional. // _t = Positive Integer: 0 to 999 inclusive. Not padded. Optional. // #y = Positive Integer: 0 to 999 inclusive. Not padded. Optional.

UPDATE

Valid file names:
File1_000_000.png File1_000_000_1.png File1_000_000#2.png File1_000_000_1#2.png File1_000_000[text].png File1_000_000[text]_1.png File1_000_000[text]#2.png File1_000_000[text]_1#2.png

The regex I've been trying is: ^(.{4}\\d_\\d{3}_\\d{3}(\\[\\w\\s]+\\])?(_\\d{1,3})?(\\#\\d{1,3})?)

This returns true for all the sample file names BUT, if I change File1_000_000[text]_1#2.png to File1_000_000[text]_#2.png by deleting the digit 1, it still returns true. The underscore is a part of the _t.

Can you qualify what you mean by "character" for "x". Digits and word characters? In the meantime, will assume you mean legal filename characters for Windows (since this is a C# question) — Neil Fenwick
– Neil Fenwick, Commented Feb 23, 2012 at 16:48
"// T = Digit: 2 to 7 inclusive." - but your example is File1. Do you mean [1-7]? — Sam Greenhalgh
– Sam Greenhalgh, Commented Feb 23, 2012 at 17:20
"x" could be any letter, digit, whitespace or legal file character. — Raheel Khan
– Raheel Khan, Commented Feb 24, 2012 at 5:59

Rich O'Kelly · Accepted Answer · 2012-02-23 16:36:27Z

2

A question on regex that doesn't involve HTML parsing, is a rarity!

Try the following:

@"^.{4}[2-7](_\d{3}){2}(\[.*?\])?(_\d{1,3})?(#\d{1,3})?\.png$"

This breaks down into:

^             Start of string
.{4}          Any character, exactly 4 times
[2-7]         A number in the range 2 - 7 once
(_\d{3}){2}   An underscore followed by 3 numbers, twice
(\[.*?\])?    An opening square bracket followed by any number of characters and closed by a square bracket 0 or 1 times
(_\d{1,3})?   An underscore followed by at least 1 and up to 3 numbers 0 or 1 times
(#\d{1,3})?   A pound (#) followed by at least 1 and up to 3 numbers 0 or 1 times
\.png$        Ending in .png

edited Feb 23, 2012 at 16:36

answered Feb 23, 2012 at 16:31

Rich O'Kelly

41.8k9 gold badges87 silver badges114 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Ry- Over a year ago

Regex tips: [\d] = \d, {0,1} = ?. Also, you need to escape the . in .png.

Raheel Khan Over a year ago

Thanks. Is there a way to extract these matched parts using the same RegEx instance? I've been manually parsing otherwise.

Rich O'Kelly Over a year ago

@RaheelKhan Yes, any portion in braces forms a capture group. Have a look at msdn.microsoft.com/en-us/library/bs2twtah.aspx for some more information on grouping.

Ry- · Accepted Answer · 2012-02-23 16:32:28Z

2

I'll just rewrite one, here:

^.{4}[2-7](_\d{3}){2}(\[[^\]]*\])?(_\d{1,3})?(\#\d{1,3})?\.png$

The problem right now is that you're not matching .png and you're not anchoring the end - the match ends prematurely. Also, you can avoid the double-escaping by prefixing your string with @:

@"^.{4}[2-7](_\d{3}){2}(\[[^\]]*\])?(_\d{1,3})?(\#\d{1,3})?\.png$"

answered Feb 23, 2012 at 16:32

Ry-♦

226k56 gold badges496 silver badges504 bronze badges

2 Comments

Raheel Khan Over a year ago

If I add braces to enclose content after start and before end like so: @"^(.{4}[2-7](_\d{3}){2}([[^]]*])?(_\d{1,3})?(\#\d{1,3})?\.png)$", does that have any implications?

Ry- Over a year ago

@RaheelKhan: No. Why would you do it, though?

Jason · Accepted Answer · 2012-02-23 16:49:17Z

1

Based on you "valid format", this will do the trick

^(?i)([a-z]{4}[2-7](_\d{3}){2}(\[.*?[^0-9]\])?(_\d{1,3}?)?(#\d{1,3}?)?\.png)$

remove (?i) to make the match case sensitive, and change [2-7] to [1-7] to make it match the files you gave (you said valid were 2-7, but your sample files are File1...)

answered Feb 23, 2012 at 16:49

Jason

3,9901 gold badge24 silver badges48 bronze badges

Comments

Sam Greenhalgh · Accepted Answer · 2012-02-23 17:35:15Z

This unit test will fail because you have specified the pattern must have an integer 2-7 at the 4th character positionm, where as File1 has 1.

[Test]
public void StackOverflow()
{
    Regex pattern = new Regex(@"^.{4}[2-7](_\d{3}){2}(\[[^\]]+\])?(_\d{1,3})?(#\d{1,3})?\.png$");

    Assert.IsTrue(pattern.IsMatch("File1_000_000.png"));
    Assert.IsTrue(pattern.IsMatch("File1_000_000_1.png"));
    Assert.IsTrue(pattern.IsMatch("File1_000_000#2.png"));
    Assert.IsTrue(pattern.IsMatch("File1_000_000_1#2.png"));
    Assert.IsTrue(pattern.IsMatch("File1_000_000[text].png"));
    Assert.IsTrue(pattern.IsMatch("File1_000_000[text]_1.png"));
    Assert.IsTrue(pattern.IsMatch("File1_000_000[text]#2.png"));
    Assert.IsTrue(pattern.IsMatch("File1_000_000[text]_1#2.png"));

    Assert.IsFalse(pattern.IsMatch("File1_000_000[text]_#2.png"));
}

Sufian Latif · Accepted Answer · 2012-02-23 17:39:18Z

0

This one should work:

^.{4}[2-7]{3}(_[0-9]{3}){2}(\[.+?\])?(_[1-9][0-9]{0,2})?(#[1-9][0-9]{0,2})?\.png$

edited Feb 23, 2012 at 17:39

answered Feb 23, 2012 at 16:39

Sufian Latif

13.4k3 gold badges36 silver badges71 bronze badges

Comments

animuson · Accepted Answer · 2012-02-23 19:34:09Z

0

@"^.{4}[2-7](_\d{3}){2}(\[.+\])?(_\d{1,3})?(\#\d{1,3})?\.png$"

edited Feb 23, 2012 at 19:34

animuson♦

54.9k28 gold badges142 silver badges150 bronze badges

answered Feb 23, 2012 at 16:45

José Mira

7517 silver badges4 bronze badges

Collectives™ on Stack Overflow

Regular expression C#

6 Answers 6

3 Comments

2 Comments

Comments

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

6 Answers 6

3 Comments

2 Comments

Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related