Matching line with 3 or more repetitions of the same date

[numberwhun]

Hello everyone!

I have a data file that contains miscellaneous information on each line. (Unfortunately, I cannot go into detail of the file layout as it is sensitive information), but I can say that on each line are dates in multiple positions. Some areas of a line contain multiple dates, strung together because there is no separation between the fields (ie: no space(s)):

ie: 05/26/200706/03/200707/24/2007

As you can see, there are 3 dates above. Some have fields following each other that are as above, only there is at least one or more spaces between the dates:

ie: 05/26/2007 06/03/2007 07/24/2007

This is where it gets a little hairy. In some files, there is an arbitrary string of dates. The string of dates would look as follows:

ie: 07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007

There may be 2, 5, 7, or more dates, all strung together, but they are all the same date through the string on that line.

I am trying to write a regex that will match this arbitrarty string of identical dates, but unfortunately, it matches any string of multiple dates. Here is what I have so far:

if($line =~ m/(\d+\/\d+\/\d+\s*){2,}/)
{
    print("Line Number $. ==>  $line \n")
}
 

Can anyone please tell me how I can match this string format:

07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007

and not any other, whether there are two dates or a dozen or so?

Regards,

Jeff

[miller]

while (<DATA>) {
    if (m{(\d{2}/\d{2}/\d{4})(?:\s*\1){2,}}) {
        print $_;
    }
}
 
__DATA__
05/26/200706/03/200707/24/2007
05/26/2007 06/03/2007 07/24/2007
07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007
 

- Miller

[numberwhun]

Miller,

WOW!! I have been at this all day and my brain is sore. Thank you! Thank you!

Now, any chance of you telling me which part of that regex says that the dates have to be identical?

Regards,

Jeff

[numberwhun]

Ok, ran this against some data I put together really quickly and I think there may still be an issue.

The dates aren't the only thing on each line. There is text before and after the dates. They are just some of the fields. I say fields as the file being examined is kind of like a flat file db where each line is a record.

The test data I just ran against is:

### Begin data###

This is a line of text.
This is another line of text.
This is the thirs line of text. 07/24/2007 07/24/2007 07/24/2007 07/24/2007
07/24/2007 07/24/2007 07/24/2007 07/24/2007
This is the fifth line.
07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007

### End Data ###

Now, when I ran the regex against this, it printed:

Line Number 6 ==> 07/24/2007 07/24/2007 07/24/2007 07/24/2007

Problem is, the data printed is from line #4.

Regards,

Jeff

[miller]

Now, any chance of you telling me which part of that regex says that the dates have to be identical?

no problem at all.

while (<DATA>) {
    if (m{
        (\d{2}/\d{2}/\d{4})   # Capture date of the format MM/DD/YYYY or DD/MM/YYYY
        (?:                   # Non-capturing group
            \s*               # Arbitrary Spacing
            \1                # Repetition of \1 date
        ){2,}                 # 2 or more copies of group.  Implies 3 or more repetitions of \1
    }xms) {
        print $_;
    }
}
 
__DATA__
05/26/200706/03/200707/24/2007
05/26/2007 06/03/2007 07/24/2007
07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007 07/24/2007
 

- Miller

[miller]

Now, when I ran the regex against this, it printed:
Line Number 6 ==> 07/24/2007 07/24/2007 07/24/2007 07/24/2007
Problem is, the data printed is from line #4.

Can't really help without knowing how you're calculating the line numbers of your data.

while (<DATA>) {
    if (m{(\d{2}/\d{2}/\d{4})(?:\s*\1){2,}}xms) {
        print "Line number $. => $_";
    }
}
 
__DATA__
This is a line of text.
This is another line of text.
This is the thirs line of text. 07/24/2007 07/24/2007 07/24/2007
07/24/2007 07/24/2007 07/24/2007 07/24/2007
This is the fifth line.
07/24/2007 07/24/2007 07/24/2007
 

Outputs:

>perl scratch.pl
Line number 3 => This is the thirs line of text. 07/24/2007 07/24/2007 07/24/2007
Line number 4 => 07/24/2007 07/24/2007 07/24/2007 07/24/2007
Line number 6 => 07/24/2007 07/24/2007 07/24/2007
 

- Miller

[numberwhun]

Believe it or not, I am letting Perl do that for me with the $. variable, which is supposed to contain the line number of the most recently read string of data.

Actually, in looking at what you did, I see you added an "xms" as options to the regex. I didn't think of that, but that seemed to work somewhat. Now it outputs the 3 lines, he he he, but all with "Line Number 6". I am so disturbed right now. I have to jet home shortly, but will post my findings later. Thanks for your help Miller!!!

Regards,

Jeff

[numberwhun]

I changed something. I was cycling through the file using a foreach loop:

foreach my $line (<FILE>)
{
    code from above
}
 

On a hunch, I changed from using the foreach to the while loop instead and it seems to work fine. Now I am curious as to why the strange behavior with the foreach. I know that with a foreach I have seen inheritance issues and that has cause some issues, but with line numbers? That's just odd. If you have any thoughts, I would be open to them.

Thanks for all your assistance!

Regards,

Jeff

[miller]

The xms option shouldn't effect your line numbering. I simply included that from my commented version of the regex.

My guess is that you're using a foreach for your file interator instead of a while loop. This would cause the entire file to be slurped the first time it was called, and therefore the $. variable would be stuck at the last line value.

Always use while loops for file loop processing. If you need filtering, then use the next command inside the while loop.

- Miller

[miller]

I changed something. I was cycling through the file using a foreach loop:

Exactly. :) See explanation above.

- M

[numberwhun]

I have said it before and I will say it again. I learn something new every day being a Perl coder, and I LOVE IT!!

Thanks again Miller!!

Regards,

Jeff

[KevinADC]

I didn't read the entire thread so I may have missed something, but this regexp looks like it need to use the string anchors ^$ to match the repeated date and only the repeated date from start to finish of the line:

if (m{^(\d{2}/\d{2}/\d{4})(?:\s*\1){2,}$}) {

otherwise it is only matching a sub string and a different date at the beginning or end will still return true .

[miller]

Yes, if you read the entire thread, you'll see that anchors are not desired. He only wishes to match a substring with the given properties.

- Miller

[KevinADC]

Yes, if you read the entire thread, you'll see that anchors are no desired. He only wishes to match a substring with the given properties.

- Miller

OK, gotcha. :)

......