It uses format strings that look much like strftime's formats, but it uses them to parse rather than format.

Hi Mark,

The DateTime project is working on unifying the chaos of perl date and time modules. You should definitely look into that before coming up with your own ad-hoc solution.

--
જલધર

right ! Murat

1. Develop Date::Manip::Lite if the original is too big.
2. Use Regexp::Common
3. Rebuild the wheel.

------
We are the carpenters and bricklayers of the Information Age.

Don't go borrowing trouble. For programmers, this means Worry only about what you need to implement.

Please remember that I'm crufty and crochety. All opinions are purely mine and all code is untested, unless otherwise specified.

I did a small date parser for very simple in Python, and it has a very simplicistic syntax. The letters Y,M and D are used to designate the numbers which should be extracted. Any other character must appear verbatim in the string, and this is validated against.

  Date( key, 'YYMMDD' )   # Date must be 6 digits
  Date( key, 'YY-MM-DD' ) # Date must match /\d\d-\d\d-\d\d/
  Date( key, 'DD.MM.YYYY' ) # Date in german format, 4 digit year
[download]

The implementation is very simple: As all strings are of fixed length, I extract the numbers from the left and right offsets of each letter (Y,M,D). After the extraction, I replace Y,M,D by "x", and then check that the string matches the string where the numbers also have been replaced by "x".

For my purposes (parsing various lists), this has proven good enough, as I don't have to deal with weekdays etc., that is, variable text except for digits.

perl -MHTTP::Daemon -MHTTP::Response -MLWP::Simple -e ' ;    # The  
$d = new HTTP::Daemon and fork and getprint $d->url and exit;#spider
($c = $d->accept())->get_request(); $c->send_response( new   #in the
HTTP::Response(200,$_,$_,qq(Just another Perl hacker\n))); ' #  web
[download]

There are the new DateTime::Format modules, but I don't see a way they can currently be used to dispatch a given date string to the right module. You could keep a list of regexen that each DateTime::Format module matches against and then send the date string to the first module that matches. You'll want to talk to the people on the datetime@perl.org mailing list first, though.

----
I wanted to explore how Perl's closures can be manipulated, and ended up creating an object system by accident.
-- Schemer

Note: All code is untested, unless otherwise stated

use strict;
sub prepareFormat {
    my $format = shift();
    my ($i, @order) = 0;
    $format =~ s/([YMDhms]+)(\?)?/ $order[$i++] = substr($1,0,1); '('.
+('\d' x length($1))."$2)"/ge;
    $format = qr/^(?:$format)$/;
    return [$format, \@order];
}

sub parseDate {
    my ($format, $date) = @_;
    my @data = ($date =~ $format->[0])
        or return;
    my %result;
    for(my $i = 0; $i <= $#data; $i++) {
        $result{$format->[1]->[$i]} ||= $data[$i];
    }
    return map $result{$_}, qw(Y M D h m s);
}

#my $format = prepareFormat ('YYYY/MM/DD');
#my $format = prepareFormat ('YYYY-DD-MM|YYYY-DD-M|YYYY-D-MM|YYYY-D-M'
+);
#my $format = prepareFormat ('YYYY-DD?-MM?');
#my $format = prepareFormat ('YYYY-DD?-MM? hh?:mm?');
#my $format = prepareFormat ('YYYY-DD?-MM? hh?:mm?(?::ss?)?');
my $format = prepareFormat ('YYYY-DD?-MM?(?: hh?:mm?(?::ss?)?)?');

while (<>) {
    chomp;
    my ($year, $month, $day, $hour, $min, $sec) = parseDate($format, $
+_)
        or print "\tNot in the right format!\n" and next;
    print "Year: $year, Month: $month, Day: $day, Hour: $hour, Min: $m
+in, Sec: $sec\n";
}
[download]

The prepareFormat() returns a reference to an array containing the regexp and the mapping between the capturing parenthesis and the parts of the date, the parseDate() uses this structure to match the date and extract the data. It then returns (year,month,day,hour,min,sec).

As you can see from the examples you may accept&parse several different datetime formats at once.

Jenda
Always code as if the guy who ends up maintaining your code will be a violent psychopath who knows where you live.
   -- Rick Osborne

Edit by castaway: Closed small tag in signature

Incredible.

This is exactly the kind of thing I was looking for. It doesn't handle AM/PM, which important in my case. Below is my hacked version that supports AM/PM by using "pp" to designate this. This creates an exceptional case in the script, because before only numbers were accepted, so instead of just \d, I'm now matching against \dPpAaMm

# test like this: perl this_script.pl '02/04/2003 06:30 PM'
use strict;
sub prepareFormat {
    my $format = shift;

    # TODO: check that only valid characters appear in the format
    # The logic should be: for any character A-Z in the format string,
    #   die if it's not one of: Y M D h m s p  

    my ($i, @order) = 0;
    my $num_chars  = 'YMDhms';
    $format =~ s/([$num_chars]+)(\?)?/ $order[$i++] = substr($1,0,1); 
+'('.('\d' x length($1))."$2)"/ge;

    # use "p" for AM/PM
    $format =~ s/pp/ $order[$i++] = substr('p',0,1); "(AM|PM)"/e;


    $format = qr/^(?:$format)$/;
    return [$format, \@order];
}

sub parseDate {
    my ($format, $date) = @_;
    my @data = ($date =~ $format->[0])
        or return;
    my %result;
    for(my $i = 0; $i <= $#data; $i++) {
        $result{$format->[1]->[$i]} ||= $data[$i];
    }

    $result{h} += 12 if ($result{p} eq 'PM' and $result{h} != 12);
    $result{h} = 0   if ($result{p} eq 'AM' and $result{h} == 12);


    return map $result{$_}, qw(Y M D h m s);
}

#my $format = prepareFormat ('YYYY/MM/DD');
#my $format = prepareFormat ('YYYY-DD-MM|YYYY-DD-M|YYYY-D-MM|YYYY-D-M'
+);
#my $format = prepareFormat ('YYYY-DD?-MM?');
#my $format = prepareFormat ('YYYY-DD?-MM? hh?:mm?');
#my $format = prepareFormat ('YYYY-DD?-MM? hh?:mm?(?::ss?)?');
#my $format = prepareFormat ('YYYY-DD?-MM?(?: hh?:mm?(?::ss?)?)?');
my $format = prepareFormat ('MM/DD/YYYY hh:mm pp');

    my ($year, $month, $day, $hour, $min, $sec) = parseDate($format, $
+ARGV[0])
        or print "\tNot in the right format!\n" and exit;
    print "Year: $year, Month: $month, Day: $day, Hour: $hour, Min: $m
+in, Sec: $sec\n";
[download]

-mark

First a tiny nit. substr('p',0,1); is better written as 'p' ;-)

Next there is a potential problem. Your code works correctly only if the AM/PM is the last thing in the string. This may seem to be an unimportant restriction, but imagine the format was 'YYYY/MM/DD hh:mm pp|DD.MM.YYYY hh:mm pp'. In this case the indexes in @order will be incorrect!

The solution is to process the 'pp' at the same time as the other characters:

use strict;
sub prepareFormat {
    my $format = shift;

    # TODO: check that only valid characters appear in the format
    # The logic should be: for any character A-Z in the format string,
    #   die if it's not one of: Y M D h m s p

    my ($i, @order) = 0;
    my $num_chars  = 'YMDhms';
    $format =~ s{([$num_chars]+|pp)(\?)?}{
        $order[$i++] = substr($1,0,1);
        if ($1 eq 'pp') {
            "(AM|PM|am|pm)"
        } else {
            '('.('\d' x length($1))."$2)"
        }
    }ge;

    $format = qr/^(?:$format)$/;
    return [$format, \@order];
}

sub parseDate {
    my ($format, $date) = @_;
    my @data = ($date =~ $format->[0])
        or return;
    my %result;
    for(my $i = 0; $i <= $#data; $i++) {
        $result{$format->[1]->[$i]} ||= $data[$i];
    }

    $result{h} += 12 if (uc($result{p}) eq 'PM' and $result{h} != 12);
    $result{h} = 0   if (uc($result{p}) eq 'AM' and $result{h} == 12);


    return map $result{$_}, qw(Y M D h m s);
}
...
[download]

Jenda
Always code as if the guy who ends up maintaining your code will be a violent psychopath who knows where you live.
   -- Rick Osborne

Edit by castaway: Closed small tag in signature