in reply to regex not working

Because your pattern 
/^(\d+)-(\d+)-(\d+) (\d+):(\d+):(\d+)$/

[download]
doesn't match your data. Particularly the milliseconds.

Update: a pattern that matches the data you show is: 

/^(\d+)-(\d+)-(\d+) (\d+):(\d+):(\d+\.\d+)$/

[download]
Update 2: a pattern that matches with optional fractional seconds is: 
/^(\d+)-(\d+)-(\d+) (\d+):(\d+):([0-9.]+)$/

[download]

Replies are listed 'Best First'.
Re^2: regex not working
by gwhite (Friar) on Jan 29, 2005 at 16:37 UTC

    If I only want up to the . and it does not always appear (different databases return different format strings for that portion of the date/time) will this:

    /^(\d+)-(\d+)-(\d+) (\d+):(\d+):(\d+)/

    [download]

    get it done? I am a regex incompetent.

    g_White
[reply]
[d/l]
      To me this last expression will not match if a . is indeed present.

      So if you don't know wether or not the milliseconds have a decimal part or not you could do : 

       /^(\d+)-(\d+)-(\d+) (\d+):(\d+):(\d+(?:\.\d+)?)$/

      [download]

      $6 will then hold the ms, with or without the optional decimal part.

[reply]
[d/l]
      What about the rough regex:

      /^.*?(\d+)\D(\d+)\D(\d+)\D(\d+)\D(\d+)\D((\d+)\.?(\d*)+)\D?/;

      which takes care to match 6 times digits delimited in some way?

      So you'll always get the:

      ($year, $month, $mday, $hour, $minutes, $seconds) = ($1, $2, $3, $4, $5, $6);
[reply]
[d/l]
[select]

      Yes, that will do exactly as you've stated. This is where you can try it yourself in a simple test case: 

      while ( chomp($_ = <DATA>) ) {
    print /^((\d+)\-(\d+)\-(\d+)\s+(\d+):(\d+):(\d+))/ ?
        "Matched on '$_': $1\n" : "Unmatched on '$_'\n";
}
__END__
1990-01-01 00:00:00.000
1990-01-01 00:00:00.000
1991-08-06 12:21:12
1991-08-07 13:41:15.65
1992-09-2110:10:10.100

      [download]
[reply]
[d/l]
      If you want only up to '.' then use $7 instead of $6.
[reply]

In Section Seekers of Perl Wisdom