in reply to
Group file per time interval
raiten:
I'd convert the times to seconds, and then do something like this:
$ cat t.pl
#!/usr/bin/perl
use strict; use warnings;
use Data::Dumper;
### Get a list of picture times, in seconds
my @times;
push @times, int 150*rand for 0 .. 30;
### Sort them
@times = sort { $a<=>$b } @times;
### Group together the pix whose time is less than MAX_TIME_DIFF secon
+ds
### apart
my $MAX_TIME_DIFF=10; # Minimum time between photos
my $MIN_GRP_SIZE=3; # Minimum "interesting" group size
my @groups;
my $cur_group = [ shift @times ];
while (@times) {
if ($$cur_group[-1]+$MAX_TIME_DIFF >= $times[0]) {
# small interval, add to current group
push @$cur_group, shift @times;
}
else {
# store last group (if interesting) and start
# a new one.
push @groups, $cur_group if @$cur_group >= $MIN_GRP_SIZE;
$cur_group = [ shift @times ];
}
}
print Dumper(\@groups);
$ perl t.pl
$VAR1 = [
[
32,
36,
39,
39,
40,
42,
48,
53,
55,
56,
57,
58,
59
],
[
72,
73,
77,
82
],
[
93,
103,
104,
105,
108,
111,
113,
113,
114,
116
]
];
...roboticus
When your only tool is a hammer, all problems look like your thumb.
