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/;

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 $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);


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);

}

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
11			my $collection; # name which will be inserted
12			my $path_add; # add additional info in path
13			my $verbose;
14			my $exclude;
15
16			#$verbose = 1;
17			my $debug = 0;
18			my $force = 0;
19
20			my $result = GetOptions(
21			"collection=s" => \$collection,
22			"path=s" => \$path_add,
23			"verbose!" => \$verbose,
24			"debug!" => \$debug,
25			"exclude=s" => \$exclude,
26			"force!" => \$force,
27			);
28
29	dpavlin	89	my ($node_url,$dir) = @ARGV;
30	dpavlin	64
31	dpavlin	89	die <<"_END_OF_USAGE_" if (! $node_url \|\| ! $dir);
32			usage: $0 http://localhost:1978/node/my_dir /path/to/directory
33
34			options:
35			--collection="name of collection"
36			--path=/path/to/add/at/end
37			--exclude=regex_to_exclude
38			--verbose
39			--force
40			--debug
41			_END_OF_USAGE_
42
43	dpavlin	64	if (! -e $dir) {
44			warn "directory $dir doesn't exist, skipping\n";
45			exit 1;
46			}
47
48			#my $basedir = $0;
49			#$basedir =~ s,/[^/]+$,/,;
50			#require "$basedir/filter.pm";
51
52	dpavlin	90	my $filter;
53			foreach my $f (qw/pdftotext pstotext/) {
54			my $w = which($f);
55			if ($f) {
56			$filter->{$f} = $w;
57			print STDERR "using $f filter at $w\n" if ($verbose);
58			}
59			}
60	dpavlin	64
61			#my $mm = new File::MMagic('/usr/share/misc/file/magic');
62			my $mm = new File::MMagic::XS();
63
64			my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');
65
66			select(STDERR); $\|=1;
67			select(STDOUT); $\|=1;
68
69	dpavlin	89	my $db = new Search::Estraier::Node;
70			$db->set_url($node_url);
71			$db->set_auth('admin', 'admin');
72	dpavlin	64
73			find({ wanted => \&file,
74			follow => 1,
75			follow_skip => 2,
76			no_chdir => 1,
77			}, $dir);
78
79
80			exit;
81
82	dpavlin	90	sub dump_contents {
83			my ($db,$contents,$mtime,$path,$size) = @_;
84	dpavlin	64
85	dpavlin	90	return unless (defined($contents)); # don't die on empty files
86	dpavlin	64
87			if ($exclude && $path =~ m/$exclude/i) {
88			print STDERR "skip: $path\n" if ($verbose);
89			return;
90			}
91
92			use bytes;
93	dpavlin	90	if (! $size) {
94			$size = length $contents;
95			}
96	dpavlin	64
97			print STDERR " [$size]" if ($verbose);
98
99			# create a document object
100	dpavlin	89	my $doc = new Search::Estraier::Document;
101	dpavlin	64
102			my $title = $1 if ($contents =~ m#<title>(.+?)</title>#is);
103
104			# chop long titles to 100 chars
105			$title = substr($title, 0, 100) . '...' if ($title && length($title) > 100);
106			# use path if no title is found
107			$title \|\|= $path;
108
109			# add attributes to the document object
110			$doc->add_attr('@uri', "file:///$path");
111			$doc->add_attr('@title', $iconv->convert($title));
112			$doc->add_attr('@size', $size);
113			$doc->add_attr('@mtime', $mtime);
114
115	dpavlin	90	if ($contents) {
116			# html2text
117			$contents =~ s#<[^>]+/*>##gs;
118			$contents =~ s#\s\s+# #gs;
119	dpavlin	64
120	dpavlin	90	$doc->add_text($iconv->convert($contents));
121			}
122			# store path
123			$doc->add_hidden_text($path);
124			# boost title
125			$doc->add_hidden_text($title);
126	dpavlin	64
127			# print $doc->dump_draft if ($verbose);
128
129			# register the document object to the database
130	dpavlin	89	$db->put_doc($doc);
131	dpavlin	64
132			}
133
134	dpavlin	90	sub filter_to_pages {
135			my ($path, $mtime, $command) = @_;
136	dpavlin	64
137			print STDERR "$path {converting}" if ($verbose);
138
139	dpavlin	90	open(F,"$command \|") \|\| die "can't open $command with '$path': $!";
140	dpavlin	64	my $html;
141			while(<F>) {
142			$html .= $_;
143			}
144			close(F);
145
146			return if (! $html);
147
148			my $file_only = $path;
149			$file_only =~ s/^.*\/([^\/]+)$/$1/g;
150
151			my ($pre_html,$pages,$post_html) = ('<html><head><title>$path :: page ##page_nr##</title></head><body><pre>',$html,'</pre></body></html>');
152
153			($pre_html,$pages,$post_html) = ($1,$2,$3) if ($html =~ m/^(<html>.+?<pre>)(.+)(<\/pre>.+?)$/si);
154
155			if ($collection) {
156			$pre_html =~ s/<title>(.+?)<\/title>/<title>$collection :: page ##page_nr##<\/title>/si;
157			} else {
158			$pre_html =~ s/<title>(.+?)<\/title>/<title>$1 :: page ##page_nr##<\/title>/si \|\|
159			$pre_html =~ s/<title><\/title>/<title>$file_only :: page ##page_nr##<\/title>/si;
160			}
161
162			# save empty entry as a placeholder
163			dump_contents($db, ' ', $mtime, "$path");
164
165			my $page_nr = 1;
166			foreach my $page (split(/\f/s,$pages)) {
167			print STDERR " $page_nr" if ($verbose);
168			my $pre_tmp = $pre_html;
169			$pre_tmp =~ s/##page_nr##/$page_nr<\/title>/s;
170			dump_contents($db, $pre_tmp . $page . $post_html, $mtime, "$path#$page_nr") if ($page !~ m/^\s*$/s);
171			$page_nr++;
172			}
173
174	dpavlin	90
175
176			}
177
178			sub file {
179
180			my $path = $_;
181			my $contents;
182
183			return if (! $force && -l $path \|\| $path =~ m#/.svn# \|\| $path =~ m/(~\|.bak)$/);
184
185			my $mtime = (stat($path))[9] \|\| -1;
186			my $mtime_db = $db->get_doc_attr_by_uri("file:///$path", '@mtime') \|\| -2;
187
188			if ($mtime == $mtime_db) {
189			print STDERR "# same: $path $mtime\n" if ($verbose);
190			return unless($force);
191	dpavlin	64	} else {
192	dpavlin	90	print STDERR "# changed: $path $mtime != $mtime_db\n" if ($debug);
193			}
194	dpavlin	64
195	dpavlin	90	# skip files on which File::MMagic::XS croaks
196			if ($path =~ m#\.au$#) {
197			warn "skipped '$path' to prevent File::MMagic::XS bug\n" if ($debug);
198			return;
199			}
200
201			my $type = $mm->checktype_filename($path);
202			$type =~ s/\s+/ /gs;
203
204			print STDERR "# $path $type\n" if ($debug);
205
206			if ($type =~ m/pdf/i) {
207			if ($filter->{pdftotext}) {
208			filter_to_pages($path, $mtime, qq( $filter->{pdftotext} -htmlmeta "$path" - ));
209			} else {
210			warn "skipping '$path', no pdftotext filter\n" if ($verbose);
211			return;
212			}
213			} elsif ($type eq 'application/postscript') {
214			if ($filter->{pstotext}) {
215			filter_to_pages($path, $mtime, qq( $filter->{pstotext} "$path" ));
216			} else {
217			warn "skipping '$path', no pstotext filter\n" if ($verbose);
218			return;
219			}
220			} else {
221
222	dpavlin	64	# return if (! -f $path \|\| ! m/\.(html*\|php\|pl\|txt\|info\|log\|text)$/i);
223	dpavlin	77	if (-f $path && $type =~ m/html/ \|\|
224	dpavlin	90	($path !~ m/\.(php\|pl\|txt\|info\|log\|text)$/io)
225	dpavlin	77	) {
226	dpavlin	90	dump_contents($db, '', $mtime, $path, -s $path);
227			return;
228	dpavlin	77	}
229	dpavlin	64
230			# skip index files
231			return if (m/index_[a-z]\.html/i \|\| m/index_symbol\.html/i);
232
233			open(F,"$path") \|\| die "can't open file: $path";
234			print STDERR "$path ($type)" if ($verbose);
235			while(<F>) {
236			$contents .= "$_";
237			}
238			$contents .= "\n\n";
239
240			#$contents = filter($contents,$collection);
241
242			# add optional components to path
243			$path .= " $path_add" if ($path_add);
244
245			dump_contents($db, $contents, $mtime, $path);
246			}
247
248			print STDERR "\n" if ($verbose);
249			# die "zero size content in '$path'" if (! $contents);
250
251			}
252