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;
#use File::MMagic;
use File::MMagic::XS qw/:compat/;

# 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 $debug = 0;
my $force = 0;

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

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

if (! -e $dir) {
        warn "directory $dir doesn't exist, skipping\n";
        exit 1;
}

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

my $pdftotext = which('pdftotext');

#my $mm = new File::MMagic('/usr/share/misc/file/magic');
my $mm = new File::MMagic::XS();

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);

        # chop long titles to 100 chars
        $title = substr($title, 0, 100) . '...' if ($title && length($title) > 100);
        # use path if no title is found
        $title ||= $path;

        # add attributes to the document object 
        $doc->add_attr('@uri', "file:///$path");
        $doc->add_attr('@title', $iconv->convert($title));
        $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 (! $force && -l $path || $path =~ m#/.svn# || $path =~ m/(~|.bak)$/);

        my $mtime = (stat($path))[9] || -1;
        my $mtime_db = $db->get_doc_attr_by_uri("file:///$path", '@mtime') || -2;

        if ($mtime == $mtime_db) {
                print STDERR "# same: $path $mtime\n" if ($verbose);
                return unless($force);
        } else {
                print STDERR "# changed: $path $mtime != $mtime_db\n" if ($debug);
        }

        # skip files on which File::MMagic::XS croaks
        return if ($path =~ m#\.au$#);

        my $type = $mm->checktype_filename($path);
        $type =~ s/\s+/ /gs;

        print STDERR "# $path $type\n" if ($debug);

        if ($pdftotext && -f $path && $type =~ 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, $mtime, "$path#$page_nr") if ($page !~ m/^\s*$/s);
                        $page_nr++;
                }

        } else {

#               return if (! -f $path || ! m/\.(html*|php|pl|txt|info|log|text)$/i);
                return unless (-f $path && $type =~ m/html/ ||
                        ($type =~ m#text# && $path =~ m/\.(php|pl|txt|info|log|text)$/io)
                );

                # 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 ($type)" if ($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, $mtime, $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	dpavlin	27	#use File::MMagic;
9			use File::MMagic::XS qw/:compat/;
10	dpavlin	5
11	dpavlin	18	# do we use Node API?
12			my $node_url;
13
14	dpavlin	5	my $collection; # name which will be inserted
15			my $path_add; # add additional info in path
16			my $verbose;
17			my $exclude;
18
19			#$verbose = 1;
20	dpavlin	28	my $debug = 0;
21	dpavlin	29	my $force = 0;
22	dpavlin	5
23			my $result = GetOptions(
24			"collection=s" => \$collection,
25			"path=s" => \$path_add,
26			"verbose!" => \$verbose,
27	dpavlin	28	"debug!" => \$debug,
28	dpavlin	5	"exclude=s" => \$exclude,
29	dpavlin	18	"node=s" => \$node_url,
30	dpavlin	29	"force!" => \$force,
31	dpavlin	5	);
32
33			my $dir = shift @ARGV \|\| die "usage: $0 [dir]";
34
35	dpavlin	31	if (! -e $dir) {
36			warn "directory $dir doesn't exist, skipping\n";
37			exit 1;
38			}
39
40	dpavlin	5	#my $basedir = $0;
41			#$basedir =~ s,/[^/]+$,/,;
42			#require "$basedir/filter.pm";
43
44			my $pdftotext = which('pdftotext');
45
46	dpavlin	27	#my $mm = new File::MMagic('/usr/share/misc/file/magic');
47			my $mm = new File::MMagic::XS();
48
49	dpavlin	5	my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');
50
51			select(STDERR); $\|=1;
52			select(STDOUT); $\|=1;
53
54			print STDERR "using $pdftotext to convert pdf into html\n" if ($pdftotext && $verbose);
55
56	dpavlin	18	my $db;
57			if ($node_url) {
58			$db = HyperEstraier::Node->new($node_url);
59			$db->set_auth('admin', 'admin');
60			} else {
61			# open the database
62			$db = HyperEstraier::Database->new();
63			$db->open('/tmp/casket', $HyperEstraier::Database::DBWRITER \| $HyperEstraier::Database::DBCREAT);
64	dpavlin	5
65	dpavlin	18	sub signal {
66			my($sig) = @_;
67			print "\nCaught a SIG$sig--syncing database and shutting down\n";
68			$db->sync();
69			exit(0);
70			}
71	dpavlin	5
72	dpavlin	18	$SIG{'INT'} = \&signal;
73			$SIG{'QUIT'} = \&signal;
74	dpavlin	5	}
75
76			find({ wanted => \&file,
77			follow => 1,
78	dpavlin	26	follow_skip => 2,
79			no_chdir => 1,
80	dpavlin	5	}, $dir);
81
82	dpavlin	18	unless ($node_url) {
83			print "--- sync\n";
84			$db->sync();
85	dpavlin	5
86	dpavlin	18	print "--- optimize...\n";
87			$db->optimize(0);
88			}
89	dpavlin	5	exit;
90
91			sub dump_contents($$$$) {
92			my ($db,$contents,$mtime,$path) = @_;
93
94			return unless ($contents); # don't die on empty files
95
96			if ($exclude && $path =~ m/$exclude/i) {
97			print STDERR "skip: $path\n" if ($verbose);
98			return;
99			}
100
101			use bytes;
102			my $size = length $contents;
103
104			print STDERR " [$size]" if ($verbose);
105
106			# create a document object
107			my $doc = HyperEstraier::Document->new;
108
109	dpavlin	29	my $title = $1 if ($contents =~ m#<title>(.+?)</title>#is);
110	dpavlin	5
111	dpavlin	27	# chop long titles to 100 chars
112			$title = substr($title, 0, 100) . '...' if ($title && length($title) > 100);
113			# use path if no title is found
114			$title \|\|= $path;
115
116	dpavlin	5	# add attributes to the document object
117			$doc->add_attr('@uri', "file:///$path");
118	dpavlin	27	$doc->add_attr('@title', $iconv->convert($title));
119	dpavlin	5	$doc->add_attr('@size', $size);
120			$doc->add_attr('@mtime', $mtime);
121
122			# html2text
123			$contents =~ s#<[^>]+/*>##gs;
124			$contents =~ s#\s\s+# #gs;
125
126			$doc->add_text($iconv->convert($contents));
127
128			# print $doc->dump_draft if ($verbose);
129
130			# register the document object to the database
131	dpavlin	18	if ($node_url) {
132			$db->put_doc($doc);
133			} else {
134			$db->put_doc($doc, $HyperEstraier::Database::PDCLEAN);
135			}
136	dpavlin	5
137			}
138
139			sub file {
140
141			my $path = $_;
142			my $contents;
143
144	dpavlin	29	return if (! $force && -l $path \|\| $path =~ m#/.svn# \|\| $path =~ m/(~\|.bak)$/);
145	dpavlin	5
146	dpavlin	31	my $mtime = (stat($path))[9] \|\| -1;
147	dpavlin	28	my $mtime_db = $db->get_doc_attr_by_uri("file:///$path", '@mtime') \|\| -2;
148
149			if ($mtime == $mtime_db) {
150			print STDERR "# same: $path $mtime\n" if ($verbose);
151	dpavlin	29	return unless($force);
152	dpavlin	28	} else {
153			print STDERR "# changed: $path $mtime != $mtime_db\n" if ($debug);
154			}
155
156			# skip files on which File::MMagic::XS croaks
157			return if ($path =~ m#\.au$#);
158
159	dpavlin	27	my $type = $mm->checktype_filename($path);
160			$type =~ s/\s+/ /gs;
161	dpavlin	5
162	dpavlin	28	print STDERR "# $path $type\n" if ($debug);
163	dpavlin	27
164			if ($pdftotext && -f $path && $type =~ m/pdf/i) {
165
166	dpavlin	5	print STDERR "$path {converting}" if ($verbose);
167
168			open(F,"$pdftotext -htmlmeta \"$path\" - \|") \|\| die "can't open $pdftotext with '$path'";
169			my $html;
170			while(<F>) {
171			# XXX why pdftotext barks if I try to use this is beyond me.
172			#$contents .= $_;
173
174			$html .= $_;
175			}
176			close(F);
177
178			return if (! $html);
179
180			my $file_only = $path;
181			$file_only =~ s/^.*\/([^\/]+)$/$1/g;
182
183			my ($pre_html,$pages,$post_html) = ('<html><head><title>$path :: page ##page_nr##</title></head><body><pre>',$html,'</pre></body></html>');
184
185			($pre_html,$pages,$post_html) = ($1,$2,$3) if ($html =~ m/^(<html>.+?<pre>)(.+)(<\/pre>.+?)$/si);
186
187			if ($collection) {
188			$pre_html =~ s/<title>(.+?)<\/title>/<title>$collection :: page ##page_nr##<\/title>/si;
189			} else {
190			$pre_html =~ s/<title>(.+?)<\/title>/<title>$1 :: page ##page_nr##<\/title>/si \|\|
191			$pre_html =~ s/<title><\/title>/<title>$file_only :: page ##page_nr##<\/title>/si;
192			}
193
194			my $page_nr = 1;
195			foreach my $page (split(/\f/s,$pages)) {
196			print STDERR " $page_nr" if ($verbose);
197			my $pre_tmp = $pre_html;
198			$pre_tmp =~ s/##page_nr##/$page_nr<\/title>/s;
199	dpavlin	30	dump_contents($db, $pre_tmp . $page . $post_html, $mtime, "$path#$page_nr") if ($page !~ m/^\s*$/s);
200	dpavlin	5	$page_nr++;
201			}
202
203			} else {
204
205	dpavlin	27	# return if (! -f $path \|\| ! m/\.(html*\|php\|pl\|txt\|info\|log\|text)$/i);
206	dpavlin	28	return unless (-f $path && $type =~ m/html/ \|\|
207			($type =~ m#text# && $path =~ m/\.(php\|pl\|txt\|info\|log\|text)$/io)
208			);
209	dpavlin	5
210			# skip index files
211			return if (m/index_[a-z]\.html/i \|\| m/index_symbol\.html/i);
212
213			open(F,"$path") \|\| die "can't open file: $path";
214	dpavlin	27	print STDERR "$path ($type)" if ($verbose);
215	dpavlin	5	while(<F>) {
216			$contents .= "$_";
217			}
218			$contents .= "\n\n";
219
220			#$contents = filter($contents,$collection);
221
222			# add optional components to path
223			$path .= " $path_add" if ($path_add);
224
225	dpavlin	28	dump_contents($db, $contents, $mtime, $path);
226	dpavlin	5	}
227
228			print STDERR "\n" if ($verbose);
229			# die "zero size content in '$path'" if (! $contents);
230
231			}
232