bfilter/trunk/bfilter.pl

#!/usr/bin/perl -w
#

use strict;

# maximum entries
my $max = 0;
# minimum letters to search by
my $min_len = shift @ARGV;
$min_len = 3 unless defined($min_len);
# if more than x elements, warn to increase min_len
my $increase_at = 500;

# name of generated index
my $headlines = 'headlines';

my $debug = 1;

sub print_file {
        my $f = shift || return;
        open(F, $f) || die "$f: $!";
        while(<F>) {
                print;
        }
        close(F);
}

print qq{
var $headlines = new Object();
};

my @part_arr;
my $last_part = '';
my $total = 0;

my $max_elements = 0;

sub escape_js {
        my $t = shift || return 'undef';
        # escape single quote and backspace
        $t =~ s/(['\\])/\\$1/g && print STDERR "ESCAPED '$t'\n";
        # quote string if not number
        $t = "'$t'" unless ($t =~ m/^\d+$/);
        return $t;
}

my @lines;

while(<STDIN>) {
        chomp;

        if (!m/\t/ || m/\t$/) {
                print STDERR "SKIP '$_': no tab\n";
                next;
        }

        # remove leading spaces (which are ignored if source list was
        # sorted using locale)
        s/^\s+//;

        push @lines, $_;
}

foreach (sort { lc($a) cmp lc($b) } @lines) {

        my @data = split(/\t+/,$_);

        my $headline = shift @data || die "need at least headline!";

        if (length($headline) < $min_len) {
                print STDERR "SKIP '$_': too short\n";
                next;
        }


        # split into min_len part and rest
        my ($part,$rest) = ( substr($headline,0,$min_len), substr($headline,$min_len) );

        # make part lowercase
        $part = lc($part);

        $last_part = $part if (! $last_part);

        # new part?
        if ($part ne $last_part) {
                print STDERR $last_part,"\t",$#part_arr+1,"\n" if ($debug && $#part_arr > $increase_at);
                $max_elements = $#part_arr if ($#part_arr > $max_elements);
                print "${headlines}[",escape_js($last_part),"] = [\n ",join(",\n ",@part_arr),"];\n" if (@part_arr);
                $total += $#part_arr;
                @part_arr = ();
                $last_part = $part;
        }
        push @part_arr, "[".escape_js($headline).",".join(",",map { escape_js($_) } @data)."]";

        # break out?
        last if ($max && $total > $max);
}

print "${headlines}[",escape_js($last_part)."] = [\n ",join(",\n ",@part_arr),"];\n" if (@part_arr);
print qq{

${headlines}.min_len = $min_len;
${headlines}.length = $total;

};

print STDERR "You have more than $increase_at elements, so you should\nincrease min_len to ",$min_len+1," or higher for performance benefit.\n" if ($max_elements > $increase_at);
1	dpavlin	1	#!/usr/bin/perl -w
2			#
3
4			use strict;
5
6			# maximum entries
7			my $max = 0;
8			# minimum letters to search by
9	dpavlin	29	my $min_len = shift @ARGV;
10			$min_len = 3 unless defined($min_len);
11	dpavlin	4	# if more than x elements, warn to increase min_len
12			my $increase_at = 500;
13	dpavlin	1
14	dpavlin	10	# name of generated index
15			my $headlines = 'headlines';
16
17	dpavlin	3	my $debug = 1;
18
19	dpavlin	1	sub print_file {
20			my $f = shift \|\| return;
21			open(F, $f) \|\| die "$f: $!";
22			while(<F>) {
23			print;
24			}
25			close(F);
26			}
27
28			print qq{
29	dpavlin	10	var $headlines = new Object();
30	dpavlin	1	};
31
32			my @part_arr;
33			my $last_part = '';
34			my $total = 0;
35
36	dpavlin	4	my $max_elements = 0;
37
38	dpavlin	25	sub escape_js {
39	dpavlin	26	my $t = shift \|\| return 'undef';
40			# escape single quote and backspace
41	dpavlin	25	$t =~ s/(['\\])/\\$1/g && print STDERR "ESCAPED '$t'\n";
42	dpavlin	26	# quote string if not number
43			$t = "'$t'" unless ($t =~ m/^\d+$/);
44	dpavlin	25	return $t;
45			}
46
47	dpavlin	35	my @lines;
48
49	dpavlin	1	while(<STDIN>) {
50			chomp;
51
52	dpavlin	9	if (!m/\t/ \|\| m/\t$/) {
53			print STDERR "SKIP '$_': no tab\n";
54			next;
55			}
56
57	dpavlin	35	# remove leading spaces (which are ignored if source list was
58			# sorted using locale)
59			s/^\s+//;
60
61			push @lines, $_;
62			}
63
64	dpavlin	36	foreach (sort { lc($a) cmp lc($b) } @lines) {
65	dpavlin	35
66	dpavlin	25	my @data = split(/\t+/,$_);
67	dpavlin	7
68	dpavlin	25	my $headline = shift @data \|\| die "need at least headline!";
69
70	dpavlin	9	if (length($headline) < $min_len) {
71			print STDERR "SKIP '$_': too short\n";
72			next;
73			}
74	dpavlin	1
75	dpavlin	9
76	dpavlin	1	# split into min_len part and rest
77	dpavlin	7	my ($part,$rest) = ( substr($headline,0,$min_len), substr($headline,$min_len) );
78	dpavlin	1
79			# make part lowercase
80			$part = lc($part);
81
82			$last_part = $part if (! $last_part);
83
84			# new part?
85			if ($part ne $last_part) {
86	dpavlin	7	print STDERR $last_part,"\t",$#part_arr+1,"\n" if ($debug && $#part_arr > $increase_at);
87	dpavlin	4	$max_elements = $#part_arr if ($#part_arr > $max_elements);
88	dpavlin	26	print "${headlines}[",escape_js($last_part),"] = [\n ",join(",\n ",@part_arr),"];\n" if (@part_arr);
89	dpavlin	1	$total += $#part_arr;
90			@part_arr = ();
91			$last_part = $part;
92			}
93	dpavlin	26	push @part_arr, "[".escape_js($headline).",".join(",",map { escape_js($_) } @data)."]";
94	dpavlin	1
95			# break out?
96			last if ($max && $total > $max);
97			}
98
99	dpavlin	26	print "${headlines}[",escape_js($last_part)."] = [\n ",join(",\n ",@part_arr),"];\n" if (@part_arr);
100	dpavlin	7	print qq{
101	dpavlin	4
102	dpavlin	10	${headlines}.min_len = $min_len;
103			${headlines}.length = $total;
104	dpavlin	7
105			};
106
107			print STDERR "You have more than $increase_at elements, so you should\nincrease min_len to ",$min_len+1," or higher for performance benefit.\n" if ($max_elements > $increase_at);