Annotation of /trunk/perl/scripts/est-spider

#!/usr/bin/perl -w
use strict;
use File::Find;
use Getopt::Long;
use File::Which;
use HyperEstraier;
use Text::Iconv;

# do we use Node API?
my $node_url;

my $collection;         # name which will be inserted
my $path_add;           # add additional info in path
my $verbose;
my $exclude;

#$verbose = 1;

my $result = GetOptions(
        "collection=s" => \$collection,
        "path=s" => \$path_add,
        "verbose!" => \$verbose,
        "debug!" => \$verbose,
        "exclude=s" => \$exclude,
        "node=s" => \$node_url,
);

my $dir = shift @ARGV || die "usage: $0 [dir]";

#my $basedir = $0;
#$basedir =~ s,/[^/]+$,/,;
#require "$basedir/filter.pm";

my $pdftotext = which('pdftotext');

my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');

select(STDERR); $|=1;
select(STDOUT); $|=1;

print STDERR "using $pdftotext to convert pdf into html\n" if ($pdftotext && $verbose);

my $db;
if ($node_url) {
        $db = HyperEstraier::Node->new($node_url);
        $db->set_auth('admin', 'admin');
} else {
        # open the database
        $db = HyperEstraier::Database->new();
        $db->open('/tmp/casket', $HyperEstraier::Database::DBWRITER | $HyperEstraier::Database::DBCREAT);

        sub signal {
                my($sig) = @_;
                print "\nCaught a SIG$sig--syncing database and shutting down\n";
                $db->sync();
                exit(0);
        }

        $SIG{'INT'}  = \&signal;
        $SIG{'QUIT'} = \&signal;
}

find({ wanted => \&file, 
        follow => 1,
        follow_skip => 2,
        no_chdir => 1,
}, $dir);

unless ($node_url) {
        print "--- sync\n";
        $db->sync();

        print "--- optimize...\n";
        $db->optimize(0);
}
exit;

sub dump_contents($$$$) {
        my ($db,$contents,$mtime,$path) = @_;

        return unless ($contents);      # don't die on empty files

        if ($exclude && $path =~ m/$exclude/i) {
                print STDERR "skip: $path\n" if ($verbose);
                return;
        }

        use bytes;
        my $size = length $contents;

        print STDERR " [$size]" if ($verbose);

        # create a document object 
        my $doc = HyperEstraier::Document->new;

        my $title = $1 if ($contents =~ m#<title>(.+)</title>#is);

        # add attributes to the document object 
        $doc->add_attr('@uri', "file:///$path");
        $doc->add_attr('@title', $title || $path);
        $doc->add_attr('@size', $size);
        $doc->add_attr('@mtime', $mtime);

        # html2text
        $contents =~ s#<[^>]+/*>##gs;
        $contents =~ s#\s\s+# #gs;

        $doc->add_text($iconv->convert($contents));

#       print $doc->dump_draft if ($verbose);

        # register the document object to the database
        if ($node_url) {
                $db->put_doc($doc);
        } else {
                $db->put_doc($doc, $HyperEstraier::Database::PDCLEAN);
        }

}

sub file {

        my $path = $_;
        my $contents;

        return if (-l $path);

        if ($pdftotext && -f $path && $path =~ m/\.pdf$/i) {

                print STDERR "$path {converting}" if ($verbose);

                open(F,"$pdftotext -htmlmeta \"$path\" - |") || die "can't open $pdftotext with '$path'";
                my $html;
                while(<F>) {
                        # XXX why pdftotext barks if I try to use this is beyond me.
                        #$contents .= $_;

                        $html .= $_;
                }
                close(F);

                return if (! $html);

                my $file_only = $path;
                $file_only =~ s/^.*\/([^\/]+)$/$1/g;

                my ($pre_html,$pages,$post_html) = ('<html><head><title>$path :: page ##page_nr##</title></head><body><pre>',$html,'</pre></body></html>');

                ($pre_html,$pages,$post_html) = ($1,$2,$3) if ($html =~ m/^(<html>.+?<pre>)(.+)(<\/pre>.+?)$/si);

                if ($collection) {
                        $pre_html =~ s/<title>(.+?)<\/title>/<title>$collection :: page ##page_nr##<\/title>/si;
                } else {
                        $pre_html =~ s/<title>(.+?)<\/title>/<title>$1 :: page ##page_nr##<\/title>/si ||
                        $pre_html =~ s/<title><\/title>/<title>$file_only :: page ##page_nr##<\/title>/si;
                }

                my $page_nr = 1;
                foreach my $page (split(/\f/s,$pages)) {
                        print STDERR " $page_nr" if ($verbose);
                        my $pre_tmp = $pre_html;
                        $pre_tmp =~ s/##page_nr##/$page_nr<\/title>/s;
                        dump_contents($db, $pre_tmp . $page . $post_html,time(), $path) if ($page !~ m/^\s*$/s);
                        $page_nr++;
                }

        } else {

                return if (! -f $path || ! m/\.(html*|php|pl|txt|info|log|text)$/i);

                # skip index files
                return if (m/index_[a-z]\.html*/i || m/index_symbol\.html*/i);

                open(F,"$path") || die "can't open file: $path";
                print STDERR "$path" if ($verbose);
                while(<F>) {
                        $contents .= "$_";
                }
                $contents .= "\n\n";

                #$contents = filter($contents,$collection);

                # add optional components to path
                $path .= " $path_add" if ($path_add);

                dump_contents($db, $contents,time(), $path);
        }

        print STDERR "\n" if ($verbose);
#       die "zero size content in '$path'" if (! $contents);

}

1	dpavlin	5	#!/usr/bin/perl -w
2			use strict;
3			use File::Find;
4			use Getopt::Long;
5			use File::Which;
6			use HyperEstraier;
7			use Text::Iconv;
8
9	dpavlin	18	# do we use Node API?
10			my $node_url;
11
12	dpavlin	5	my $collection; # name which will be inserted
13			my $path_add; # add additional info in path
14			my $verbose;
15			my $exclude;
16
17			#$verbose = 1;
18
19			my $result = GetOptions(
20			"collection=s" => \$collection,
21			"path=s" => \$path_add,
22			"verbose!" => \$verbose,
23			"debug!" => \$verbose,
24			"exclude=s" => \$exclude,
25	dpavlin	18	"node=s" => \$node_url,
26	dpavlin	5	);
27
28			my $dir = shift @ARGV \|\| die "usage: $0 [dir]";
29
30			#my $basedir = $0;
31			#$basedir =~ s,/[^/]+$,/,;
32			#require "$basedir/filter.pm";
33
34			my $pdftotext = which('pdftotext');
35
36			my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');
37
38			select(STDERR); $\|=1;
39			select(STDOUT); $\|=1;
40
41			print STDERR "using $pdftotext to convert pdf into html\n" if ($pdftotext && $verbose);
42
43	dpavlin	18	my $db;
44			if ($node_url) {
45			$db = HyperEstraier::Node->new($node_url);
46			$db->set_auth('admin', 'admin');
47			} else {
48			# open the database
49			$db = HyperEstraier::Database->new();
50			$db->open('/tmp/casket', $HyperEstraier::Database::DBWRITER \| $HyperEstraier::Database::DBCREAT);
51	dpavlin	5
52	dpavlin	18	sub signal {
53			my($sig) = @_;
54			print "\nCaught a SIG$sig--syncing database and shutting down\n";
55			$db->sync();
56			exit(0);
57			}
58	dpavlin	5
59	dpavlin	18	$SIG{'INT'} = \&signal;
60			$SIG{'QUIT'} = \&signal;
61	dpavlin	5	}
62
63			find({ wanted => \&file,
64			follow => 1,
65	dpavlin	26	follow_skip => 2,
66			no_chdir => 1,
67	dpavlin	5	}, $dir);
68
69	dpavlin	18	unless ($node_url) {
70			print "--- sync\n";
71			$db->sync();
72	dpavlin	5
73	dpavlin	18	print "--- optimize...\n";
74			$db->optimize(0);
75			}
76	dpavlin	5	exit;
77
78			sub dump_contents($$$$) {
79			my ($db,$contents,$mtime,$path) = @_;
80
81			return unless ($contents); # don't die on empty files
82
83			if ($exclude && $path =~ m/$exclude/i) {
84			print STDERR "skip: $path\n" if ($verbose);
85			return;
86			}
87
88			use bytes;
89			my $size = length $contents;
90
91			print STDERR " [$size]" if ($verbose);
92
93			# create a document object
94			my $doc = HyperEstraier::Document->new;
95
96			my $title = $1 if ($contents =~ m#<title>(.+)</title>#is);
97
98			# add attributes to the document object
99			$doc->add_attr('@uri', "file:///$path");
100			$doc->add_attr('@title', $title \|\| $path);
101			$doc->add_attr('@size', $size);
102			$doc->add_attr('@mtime', $mtime);
103
104			# html2text
105			$contents =~ s#<[^>]+/*>##gs;
106			$contents =~ s#\s\s+# #gs;
107
108			$doc->add_text($iconv->convert($contents));
109
110			# print $doc->dump_draft if ($verbose);
111
112			# register the document object to the database
113	dpavlin	18	if ($node_url) {
114			$db->put_doc($doc);
115			} else {
116			$db->put_doc($doc, $HyperEstraier::Database::PDCLEAN);
117			}
118	dpavlin	5
119			}
120
121			sub file {
122
123			my $path = $_;
124			my $contents;
125
126			return if (-l $path);
127
128			if ($pdftotext && -f $path && $path =~ m/\.pdf$/i) {
129
130			print STDERR "$path {converting}" if ($verbose);
131
132			open(F,"$pdftotext -htmlmeta \"$path\" - \|") \|\| die "can't open $pdftotext with '$path'";
133			my $html;
134			while(<F>) {
135			# XXX why pdftotext barks if I try to use this is beyond me.
136			#$contents .= $_;
137
138			$html .= $_;
139			}
140			close(F);
141
142			return if (! $html);
143
144			my $file_only = $path;
145			$file_only =~ s/^.*\/([^\/]+)$/$1/g;
146
147			my ($pre_html,$pages,$post_html) = ('<html><head><title>$path :: page ##page_nr##</title></head><body><pre>',$html,'</pre></body></html>');
148
149			($pre_html,$pages,$post_html) = ($1,$2,$3) if ($html =~ m/^(<html>.+?<pre>)(.+)(<\/pre>.+?)$/si);
150
151			if ($collection) {
152			$pre_html =~ s/<title>(.+?)<\/title>/<title>$collection :: page ##page_nr##<\/title>/si;
153			} else {
154			$pre_html =~ s/<title>(.+?)<\/title>/<title>$1 :: page ##page_nr##<\/title>/si \|\|
155			$pre_html =~ s/<title><\/title>/<title>$file_only :: page ##page_nr##<\/title>/si;
156			}
157
158			my $page_nr = 1;
159			foreach my $page (split(/\f/s,$pages)) {
160			print STDERR " $page_nr" if ($verbose);
161			my $pre_tmp = $pre_html;
162			$pre_tmp =~ s/##page_nr##/$page_nr<\/title>/s;
163			dump_contents($db, $pre_tmp . $page . $post_html,time(), $path) if ($page !~ m/^\s*$/s);
164			$page_nr++;
165			}
166
167			} else {
168
169			return if (! -f $path \|\| ! m/\.(html*\|php\|pl\|txt\|info\|log\|text)$/i);
170
171			# skip index files
172			return if (m/index_[a-z]\.html/i \|\| m/index_symbol\.html/i);
173
174			open(F,"$path") \|\| die "can't open file: $path";
175			print STDERR "$path" if ($verbose);
176			while(<F>) {
177			$contents .= "$_";
178			}
179			$contents .= "\n\n";
180
181			#$contents = filter($contents,$collection);
182
183			# add optional components to path
184			$path .= " $path_add" if ($path_add);
185
186			dump_contents($db, $contents,time(), $path);
187			}
188
189			print STDERR "\n" if ($verbose);
190			# die "zero size content in '$path'" if (! $contents);
191
192			}
193