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 $all = 0;

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

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
        --all   save placeholders for all files
_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(
        url => $node_url,
        user => 'admin',
        passwd => 'admin',
        croak_on_error => 1,
);

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 $doc->dump_draft if ($debug);

        # 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# || $path =~ m/(~|.bak|.gif)$/);

        my $mtime = (stat($path))[9] || -1;
        my $mtime_db = eval { $db->get_doc_attr_by_uri("file:///$path", '@mtime') };
        $mtime_db ||= -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 $path) if ($all);
                        return;
                }

                # skip index files
                return if ($path =~ m/index_(?:[a-z]+|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	dpavlin	119	my $all = 0;
21	dpavlin	64
22			my $result = GetOptions(
23			"collection=s" => \$collection,
24			"path=s" => \$path_add,
25			"verbose!" => \$verbose,
26			"debug!" => \$debug,
27			"exclude=s" => \$exclude,
28			"force!" => \$force,
29	dpavlin	119	"all!" => \$all,
30	dpavlin	64	);
31
32	dpavlin	89	my ($node_url,$dir) = @ARGV;
33	dpavlin	64
34	dpavlin	89	die <<"_END_OF_USAGE_" if (! $node_url \|\| ! $dir);
35			usage: $0 http://localhost:1978/node/my_dir /path/to/directory
36
37			options:
38			--collection="name of collection"
39			--path=/path/to/add/at/end
40			--exclude=regex_to_exclude
41			--verbose
42			--force
43			--debug
44	dpavlin	119	--all save placeholders for all files
45	dpavlin	89	_END_OF_USAGE_
46
47	dpavlin	64	if (! -e $dir) {
48			warn "directory $dir doesn't exist, skipping\n";
49			exit 1;
50			}
51
52			#my $basedir = $0;
53			#$basedir =~ s,/[^/]+$,/,;
54			#require "$basedir/filter.pm";
55
56	dpavlin	118	my $docs = 0;
57			my $start_t = time();
58
59	dpavlin	90	my $filter;
60			foreach my $f (qw/pdftotext pstotext/) {
61			my $w = which($f);
62			if ($f) {
63			$filter->{$f} = $w;
64			print STDERR "using $f filter at $w\n" if ($verbose);
65			}
66			}
67	dpavlin	64
68			#my $mm = new File::MMagic('/usr/share/misc/file/magic');
69			my $mm = new File::MMagic::XS();
70
71			my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');
72
73			select(STDERR); $\|=1;
74			select(STDOUT); $\|=1;
75
76	dpavlin	119	my $db = new Search::Estraier::Node(
77			url => $node_url,
78			user => 'admin',
79			passwd => 'admin',
80			croak_on_error => 1,
81			);
82	dpavlin	64
83			find({ wanted => \&file,
84			follow => 1,
85			follow_skip => 2,
86			no_chdir => 1,
87			}, $dir);
88
89	dpavlin	118	my $dur = (time() - $start_t) \|\| 1;
90			printf STDERR "%d documents in %.2fs [%.2f docs/s]\n", $docs, $dur, ($docs / $dur);
91	dpavlin	64
92			exit;
93
94	dpavlin	90	sub dump_contents {
95			my ($db,$contents,$mtime,$path,$size) = @_;
96	dpavlin	64
97	dpavlin	90	return unless (defined($contents)); # don't die on empty files
98	dpavlin	64
99			if ($exclude && $path =~ m/$exclude/i) {
100			print STDERR "skip: $path\n" if ($verbose);
101			return;
102			}
103
104			use bytes;
105	dpavlin	90	if (! $size) {
106			$size = length $contents;
107			}
108	dpavlin	64
109			print STDERR " [$size]" if ($verbose);
110
111			# create a document object
112	dpavlin	89	my $doc = new Search::Estraier::Document;
113	dpavlin	64
114			my $title = $1 if ($contents =~ m#<title>(.+?)</title>#is);
115
116			# chop long titles to 100 chars
117			$title = substr($title, 0, 100) . '...' if ($title && length($title) > 100);
118			# use path if no title is found
119			$title \|\|= $path;
120
121			# add attributes to the document object
122			$doc->add_attr('@uri', "file:///$path");
123			$doc->add_attr('@title', $iconv->convert($title));
124			$doc->add_attr('@size', $size);
125			$doc->add_attr('@mtime', $mtime);
126
127	dpavlin	90	if ($contents) {
128			# html2text
129			$contents =~ s#<[^>]+/*>##gs;
130			$contents =~ s#\s\s+# #gs;
131	dpavlin	64
132	dpavlin	90	$doc->add_text($iconv->convert($contents));
133			}
134			# store path
135			$doc->add_hidden_text($path);
136			# boost title
137			$doc->add_hidden_text($title);
138	dpavlin	64
139	dpavlin	120	print $doc->dump_draft if ($debug);
140	dpavlin	64
141			# register the document object to the database
142	dpavlin	89	$db->put_doc($doc);
143	dpavlin	64
144	dpavlin	118	$docs++;
145
146	dpavlin	64	}
147
148	dpavlin	90	sub filter_to_pages {
149			my ($path, $mtime, $command) = @_;
150	dpavlin	64
151			print STDERR "$path {converting}" if ($verbose);
152
153	dpavlin	90	open(F,"$command \|") \|\| die "can't open $command with '$path': $!";
154	dpavlin	64	my $html;
155			while(<F>) {
156			$html .= $_;
157			}
158			close(F);
159
160			return if (! $html);
161
162			my $file_only = $path;
163			$file_only =~ s/^.*\/([^\/]+)$/$1/g;
164
165			my ($pre_html,$pages,$post_html) = ('<html><head><title>$path :: page ##page_nr##</title></head><body><pre>',$html,'</pre></body></html>');
166
167			($pre_html,$pages,$post_html) = ($1,$2,$3) if ($html =~ m/^(<html>.+?<pre>)(.+)(<\/pre>.+?)$/si);
168
169			if ($collection) {
170			$pre_html =~ s/<title>(.+?)<\/title>/<title>$collection :: page ##page_nr##<\/title>/si;
171			} else {
172			$pre_html =~ s/<title>(.+?)<\/title>/<title>$1 :: page ##page_nr##<\/title>/si \|\|
173			$pre_html =~ s/<title><\/title>/<title>$file_only :: page ##page_nr##<\/title>/si;
174			}
175
176			# save empty entry as a placeholder
177			dump_contents($db, ' ', $mtime, "$path");
178
179			my $page_nr = 1;
180			foreach my $page (split(/\f/s,$pages)) {
181			print STDERR " $page_nr" if ($verbose);
182			my $pre_tmp = $pre_html;
183			$pre_tmp =~ s/##page_nr##/$page_nr<\/title>/s;
184			dump_contents($db, $pre_tmp . $page . $post_html, $mtime, "$path#$page_nr") if ($page !~ m/^\s*$/s);
185			$page_nr++;
186			}
187
188	dpavlin	90
189
190			}
191
192			sub file {
193
194			my $path = $_;
195			my $contents;
196
197	dpavlin	119	return if (! $force && -l $path \|\| $path =~ m#/.svn# \|\| $path =~ m/(~\|.bak\|.gif)$/);
198	dpavlin	90
199			my $mtime = (stat($path))[9] \|\| -1;
200	dpavlin	119	my $mtime_db = eval { $db->get_doc_attr_by_uri("file:///$path", '@mtime') };
201			$mtime_db \|\|= -2;
202	dpavlin	90
203			if ($mtime == $mtime_db) {
204			print STDERR "# same: $path $mtime\n" if ($verbose);
205			return unless($force);
206	dpavlin	64	} else {
207	dpavlin	90	print STDERR "# changed: $path $mtime != $mtime_db\n" if ($debug);
208			}
209	dpavlin	64
210	dpavlin	90	# skip files on which File::MMagic::XS croaks
211			if ($path =~ m#\.au$#) {
212			warn "skipped '$path' to prevent File::MMagic::XS bug\n" if ($debug);
213			return;
214			}
215
216			my $type = $mm->checktype_filename($path);
217			$type =~ s/\s+/ /gs;
218
219			print STDERR "# $path $type\n" if ($debug);
220
221			if ($type =~ m/pdf/i) {
222			if ($filter->{pdftotext}) {
223			filter_to_pages($path, $mtime, qq( $filter->{pdftotext} -htmlmeta "$path" - ));
224			} else {
225			warn "skipping '$path', no pdftotext filter\n" if ($verbose);
226			return;
227			}
228			} elsif ($type eq 'application/postscript') {
229			if ($filter->{pstotext}) {
230			filter_to_pages($path, $mtime, qq( $filter->{pstotext} "$path" ));
231			} else {
232			warn "skipping '$path', no pstotext filter\n" if ($verbose);
233			return;
234			}
235			} else {
236
237	dpavlin	64	# return if (! -f $path \|\| ! m/\.(html*\|php\|pl\|txt\|info\|log\|text)$/i);
238	dpavlin	119	if (-f $path &&
239			$type !~ m/html/ &&
240			$path !~ m/\.(php\|pl\|txt\|info\|log\|text)$/io
241	dpavlin	77	) {
242	dpavlin	119	dump_contents($db, '', $mtime, $path, -s $path) if ($all);
243	dpavlin	90	return;
244	dpavlin	77	}
245	dpavlin	64
246			# skip index files
247	dpavlin	119	return if ($path =~ m/index_(?:[a-z]+\|symbol)\.html*/i);
248	dpavlin	64
249			open(F,"$path") \|\| die "can't open file: $path";
250			print STDERR "$path ($type)" if ($verbose);
251			while(<F>) {
252			$contents .= "$_";
253			}
254			$contents .= "\n\n";
255
256			#$contents = filter($contents,$collection);
257
258			# add optional components to path
259			$path .= " $path_add" if ($path_add);
260
261			dump_contents($db, $contents, $mtime, $path);
262			}
263
264			print STDERR "\n" if ($verbose);
265			# die "zero size content in '$path'" if (! $contents);
266
267			}
268