Annotation of /trunk/scripts/est-spider

#!/usr/bin/perl -w
use strict;
use File::Find;
use Getopt::Long;
use File::Which;
use Search::Estraier;
use Text::Iconv;
#use File::MMagic;
use File::MMagic::XS qw/:compat/;
use Time::HiRes qw/time/;

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,
        "force!" => \$force,
);

my ($node_url,$dir) = @ARGV;

die <<"_END_OF_USAGE_" if (! $node_url || ! $dir);
usage: $0 http://localhost:1978/node/my_dir /path/to/directory

options:
        --collection="name of collection"
        --path=/path/to/add/at/end
        --exclude=regex_to_exclude
        --verbose
        --force
        --debug
_END_OF_USAGE_

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

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

my $docs = 0;
my $start_t = time();

my $filter;
foreach my $f (qw/pdftotext pstotext/) {
        my $w = which($f);
        if ($f) {
                $filter->{$f} = $w;
                print STDERR "using $f filter at $w\n" if ($verbose);
        }
}

#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;

my $db = new Search::Estraier::Node;
$db->set_url($node_url);
$db->set_auth('admin', 'admin');

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

my $dur = (time() - $start_t) || 1;
printf STDERR "%d documents in %.2fs [%.2f docs/s]\n", $docs, $dur, ($docs / $dur);

exit;

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

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

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

        use bytes;
        if (! $size) {
                $size = length $contents;
        }

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

        # create a document object 
        my $doc = new Search::Estraier::Document;

        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);

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

                $doc->add_text($iconv->convert($contents));
        }
        # store path
        $doc->add_hidden_text($path);
        # boost title
        $doc->add_hidden_text($title);

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

        # register the document object to the database
        $db->put_doc($doc);

        $docs++;

}

sub filter_to_pages {
                my ($path, $mtime, $command) = @_;

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

                open(F,"$command |") || die "can't open $command with '$path': $!";
                my $html;
                while(<F>) {
                        $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;
                }

                # save empty entry as a placeholder
                dump_contents($db, ' ', $mtime, "$path");

                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++;
                }


}

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
        if ($path =~ m#\.au$#) {
                warn "skipped '$path' to prevent File::MMagic::XS bug\n" if ($debug);
                return;
        }

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

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

        if ($type =~ m/pdf/i) {
                if ($filter->{pdftotext}) {
                        filter_to_pages($path, $mtime, qq( $filter->{pdftotext} -htmlmeta "$path" - ));
                } else {
                        warn "skipping '$path', no pdftotext filter\n" if ($verbose);
                        return;
                }
        } elsif ($type eq 'application/postscript') {
                if ($filter->{pstotext}) {
                        filter_to_pages($path, $mtime, qq( $filter->{pstotext} "$path" ));
                } else {
                        warn "skipping '$path', no pstotext filter\n" if ($verbose);
                        return;
                }
        } else {

#               return if (! -f $path || ! m/\.(html*|php|pl|txt|info|log|text)$/i);
                if (-f $path && $type =~ m/html/ ||
                        ($path !~ m/\.(php|pl|txt|info|log|text)$/io)
                ) {
                        dump_contents($db, '', $mtime, $path, -s $path);
                        return;
                }

                # 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	64	#!/usr/bin/perl -w
2			use strict;
3			use File::Find;
4			use Getopt::Long;
5			use File::Which;
6			use Search::Estraier;
7			use Text::Iconv;
8			#use File::MMagic;
9			use File::MMagic::XS qw/:compat/;
10	dpavlin	118	use Time::HiRes qw/time/;
11	dpavlin	64
12			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			my $debug = 0;
19			my $force = 0;
20
21			my $result = GetOptions(
22			"collection=s" => \$collection,
23			"path=s" => \$path_add,
24			"verbose!" => \$verbose,
25			"debug!" => \$debug,
26			"exclude=s" => \$exclude,
27			"force!" => \$force,
28			);
29
30	dpavlin	89	my ($node_url,$dir) = @ARGV;
31	dpavlin	64
32	dpavlin	89	die <<"_END_OF_USAGE_" if (! $node_url \|\| ! $dir);
33			usage: $0 http://localhost:1978/node/my_dir /path/to/directory
34
35			options:
36			--collection="name of collection"
37			--path=/path/to/add/at/end
38			--exclude=regex_to_exclude
39			--verbose
40			--force
41			--debug
42			_END_OF_USAGE_
43
44	dpavlin	64	if (! -e $dir) {
45			warn "directory $dir doesn't exist, skipping\n";
46			exit 1;
47			}
48
49			#my $basedir = $0;
50			#$basedir =~ s,/[^/]+$,/,;
51			#require "$basedir/filter.pm";
52
53	dpavlin	118	my $docs = 0;
54			my $start_t = time();
55
56	dpavlin	90	my $filter;
57			foreach my $f (qw/pdftotext pstotext/) {
58			my $w = which($f);
59			if ($f) {
60			$filter->{$f} = $w;
61			print STDERR "using $f filter at $w\n" if ($verbose);
62			}
63			}
64	dpavlin	64
65			#my $mm = new File::MMagic('/usr/share/misc/file/magic');
66			my $mm = new File::MMagic::XS();
67
68			my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');
69
70			select(STDERR); $\|=1;
71			select(STDOUT); $\|=1;
72
73	dpavlin	89	my $db = new Search::Estraier::Node;
74			$db->set_url($node_url);
75			$db->set_auth('admin', 'admin');
76	dpavlin	64
77			find({ wanted => \&file,
78			follow => 1,
79			follow_skip => 2,
80			no_chdir => 1,
81			}, $dir);
82
83	dpavlin	118	my $dur = (time() - $start_t) \|\| 1;
84			printf STDERR "%d documents in %.2fs [%.2f docs/s]\n", $docs, $dur, ($docs / $dur);
85	dpavlin	64
86			exit;
87
88	dpavlin	90	sub dump_contents {
89			my ($db,$contents,$mtime,$path,$size) = @_;
90	dpavlin	64
91	dpavlin	90	return unless (defined($contents)); # don't die on empty files
92	dpavlin	64
93			if ($exclude && $path =~ m/$exclude/i) {
94			print STDERR "skip: $path\n" if ($verbose);
95			return;
96			}
97
98			use bytes;
99	dpavlin	90	if (! $size) {
100			$size = length $contents;
101			}
102	dpavlin	64
103			print STDERR " [$size]" if ($verbose);
104
105			# create a document object
106	dpavlin	89	my $doc = new Search::Estraier::Document;
107	dpavlin	64
108			my $title = $1 if ($contents =~ m#<title>(.+?)</title>#is);
109
110			# chop long titles to 100 chars
111			$title = substr($title, 0, 100) . '...' if ($title && length($title) > 100);
112			# use path if no title is found
113			$title \|\|= $path;
114
115			# add attributes to the document object
116			$doc->add_attr('@uri', "file:///$path");
117			$doc->add_attr('@title', $iconv->convert($title));
118			$doc->add_attr('@size', $size);
119			$doc->add_attr('@mtime', $mtime);
120
121	dpavlin	90	if ($contents) {
122			# html2text
123			$contents =~ s#<[^>]+/*>##gs;
124			$contents =~ s#\s\s+# #gs;
125	dpavlin	64
126	dpavlin	90	$doc->add_text($iconv->convert($contents));
127			}
128			# store path
129			$doc->add_hidden_text($path);
130			# boost title
131			$doc->add_hidden_text($title);
132	dpavlin	64
133			# print $doc->dump_draft if ($verbose);
134
135			# register the document object to the database
136	dpavlin	89	$db->put_doc($doc);
137	dpavlin	64
138	dpavlin	118	$docs++;
139
140	dpavlin	64	}
141
142	dpavlin	90	sub filter_to_pages {
143			my ($path, $mtime, $command) = @_;
144	dpavlin	64
145			print STDERR "$path {converting}" if ($verbose);
146
147	dpavlin	90	open(F,"$command \|") \|\| die "can't open $command with '$path': $!";
148	dpavlin	64	my $html;
149			while(<F>) {
150			$html .= $_;
151			}
152			close(F);
153
154			return if (! $html);
155
156			my $file_only = $path;
157			$file_only =~ s/^.*\/([^\/]+)$/$1/g;
158
159			my ($pre_html,$pages,$post_html) = ('<html><head><title>$path :: page ##page_nr##</title></head><body><pre>',$html,'</pre></body></html>');
160
161			($pre_html,$pages,$post_html) = ($1,$2,$3) if ($html =~ m/^(<html>.+?<pre>)(.+)(<\/pre>.+?)$/si);
162
163			if ($collection) {
164			$pre_html =~ s/<title>(.+?)<\/title>/<title>$collection :: page ##page_nr##<\/title>/si;
165			} else {
166			$pre_html =~ s/<title>(.+?)<\/title>/<title>$1 :: page ##page_nr##<\/title>/si \|\|
167			$pre_html =~ s/<title><\/title>/<title>$file_only :: page ##page_nr##<\/title>/si;
168			}
169
170			# save empty entry as a placeholder
171			dump_contents($db, ' ', $mtime, "$path");
172
173			my $page_nr = 1;
174			foreach my $page (split(/\f/s,$pages)) {
175			print STDERR " $page_nr" if ($verbose);
176			my $pre_tmp = $pre_html;
177			$pre_tmp =~ s/##page_nr##/$page_nr<\/title>/s;
178			dump_contents($db, $pre_tmp . $page . $post_html, $mtime, "$path#$page_nr") if ($page !~ m/^\s*$/s);
179			$page_nr++;
180			}
181
182	dpavlin	90
183
184			}
185
186			sub file {
187
188			my $path = $_;
189			my $contents;
190
191			return if (! $force && -l $path \|\| $path =~ m#/.svn# \|\| $path =~ m/(~\|.bak)$/);
192
193			my $mtime = (stat($path))[9] \|\| -1;
194			my $mtime_db = $db->get_doc_attr_by_uri("file:///$path", '@mtime') \|\| -2;
195
196			if ($mtime == $mtime_db) {
197			print STDERR "# same: $path $mtime\n" if ($verbose);
198			return unless($force);
199	dpavlin	64	} else {
200	dpavlin	90	print STDERR "# changed: $path $mtime != $mtime_db\n" if ($debug);
201			}
202	dpavlin	64
203	dpavlin	90	# skip files on which File::MMagic::XS croaks
204			if ($path =~ m#\.au$#) {
205			warn "skipped '$path' to prevent File::MMagic::XS bug\n" if ($debug);
206			return;
207			}
208
209			my $type = $mm->checktype_filename($path);
210			$type =~ s/\s+/ /gs;
211
212			print STDERR "# $path $type\n" if ($debug);
213
214			if ($type =~ m/pdf/i) {
215			if ($filter->{pdftotext}) {
216			filter_to_pages($path, $mtime, qq( $filter->{pdftotext} -htmlmeta "$path" - ));
217			} else {
218			warn "skipping '$path', no pdftotext filter\n" if ($verbose);
219			return;
220			}
221			} elsif ($type eq 'application/postscript') {
222			if ($filter->{pstotext}) {
223			filter_to_pages($path, $mtime, qq( $filter->{pstotext} "$path" ));
224			} else {
225			warn "skipping '$path', no pstotext filter\n" if ($verbose);
226			return;
227			}
228			} else {
229
230	dpavlin	64	# return if (! -f $path \|\| ! m/\.(html*\|php\|pl\|txt\|info\|log\|text)$/i);
231	dpavlin	77	if (-f $path && $type =~ m/html/ \|\|
232	dpavlin	90	($path !~ m/\.(php\|pl\|txt\|info\|log\|text)$/io)
233	dpavlin	77	) {
234	dpavlin	90	dump_contents($db, '', $mtime, $path, -s $path);
235			return;
236	dpavlin	77	}
237	dpavlin	64
238			# skip index files
239			return if (m/index_[a-z]\.html/i \|\| m/index_symbol\.html/i);
240
241			open(F,"$path") \|\| die "can't open file: $path";
242			print STDERR "$path ($type)" if ($verbose);
243			while(<F>) {
244			$contents .= "$_";
245			}
246			$contents .= "\n\n";
247
248			#$contents = filter($contents,$collection);
249
250			# add optional components to path
251			$path .= " $path_add" if ($path_add);
252
253			dump_contents($db, $contents, $mtime, $path);
254			}
255
256			print STDERR "\n" if ($verbose);
257			# die "zero size content in '$path'" if (! $contents);
258
259			}
260