1: #!/usr/bin/perl
2: # The LearningOnline Network
3: # searchcat.pl "Search Catalog" batch script
4: #
5: # $Id: searchcat.pl,v 1.19 2002/07/01 18:23:00 matthew Exp $
6: #
7: # Copyright Michigan State University Board of Trustees
8: #
9: # This file is part of the LearningOnline Network with CAPA (LON-CAPA).
10: #
11: # LON-CAPA is free software; you can redistribute it and/or modify
12: # it under the terms of the GNU General Public License as published by
13: # the Free Software Foundation; either version 2 of the License, or
14: # (at your option) any later version.
15: #
16: # LON-CAPA is distributed in the hope that it will be useful,
17: # but WITHOUT ANY WARRANTY; without even the implied warranty of
18: # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19: # GNU General Public License for more details.
20: #
21: # You should have received a copy of the GNU General Public License
22: # along with LON-CAPA; if not, write to the Free Software
23: # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
24: #
25: # /home/httpd/html/adm/gpl.txt
26: #
27: # http://www.lon-capa.org/
28: #
29: # YEAR=2001
30: # 04/14/2001, 04/16/2001 Scott Harrison
31: #
32: # YEAR=2002
33: # 05/11/2002 Scott Harrison
34: #
35: ###
36:
37: # This script goes through a LON-CAPA resource
38: # directory and gathers metadata.
39: # The metadata is entered into a SQL database.
40:
41: use lib '/home/httpd/lib/perl/';
42: use LONCAPA::Configuration;
43:
44: use IO::File;
45: use HTML::TokeParser;
46: use DBI;
47:
48: my @metalist;
49: # ----------------- Code to enable 'find' subroutine listing of the .meta files
50: require "find.pl";
51: sub wanted {
52: (($dev,$ino,$mode,$nlink,$uid,$gid) = lstat($_)) &&
53: -f _ &&
54: /^.*\.meta$/ && !/^.+\.\d+\.[^\.]+\.meta$/ &&
55: push(@metalist,"$dir/$_");
56: }
57:
58: # --------------- Read loncapa_apache.conf and loncapa.conf and get variables
59: my $perlvarref=LONCAPA::Configuration::read_conf('loncapa_apache.conf',
60: 'loncapa.conf');
61: my %perlvar=%{$perlvarref};
62: undef $perlvarref; # remove since sensitive and not needed
63: delete $perlvar{'lonReceipt'}; # remove since sensitive and not needed
64:
65: # ------------------------------------- Only run if machine is a library server
66: exit unless $perlvar{'lonRole'} eq 'library';
67:
68: my $dbh;
69: # ------------------------------------- Make sure that database can be accessed
70: {
71: unless (
72: $dbh = DBI->connect("DBI:mysql:loncapa","www",$perlvar{'lonSqlAccess'},{ RaiseError =>0,PrintError=>0})
73: ) {
74: print "Cannot connect to database!\n";
75: exit;
76: }
77: my $make_metadata_table = "CREATE TABLE IF NOT EXISTS metadata (".
78: "title TEXT, author TEXT, subject TEXT, url TEXT, keywords TEXT, ".
79: "version TEXT, notes TEXT, abstract TEXT, mime TEXT, language TEXT, ".
80: "creationdate DATETIME, lastrevisiondate DATETIME, owner TEXT, ".
81: "copyright TEXT, FULLTEXT idx_title (title), ".
82: "FULLTEXT idx_author (author), FULLTEXT idx_subject (subject), ".
83: "FULLTEXT idx_url (url), FULLTEXT idx_keywords (keywords), ".
84: "FULLTEXT idx_version (version), FULLTEXT idx_notes (notes), ".
85: "FULLTEXT idx_abstract (abstract), FULLTEXT idx_mime (mime), ".
86: "FULLTEXT idx_language (language), FULLTEXT idx_owner (owner), ".
87: "FULLTEXT idx_copyright (copyright)) TYPE=MYISAM";
88: # It would sure be nice to have some logging mechanism.
89: $dbh->do($make_metadata_table);
90: }
91:
92: # ------------------------------------------------------------- get .meta files
93: opendir(RESOURCES,"$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}");
94: my @homeusers=grep
95: {&ishome("$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}/$_")}
96: grep {!/^\.\.?$/} readdir(RESOURCES);
97: closedir RESOURCES;
98: foreach my $user (@homeusers) {
99: &find("$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}/$user");
100: }
101:
102: # -- process each file to get metadata and put into search catalog SQL database
103: # Also, check to see if already there.
104: # I could just delete (without searching first), but this works for now.
105: foreach my $m (@metalist) {
106: my $ref=&metadata($m);
107: my $m2='/res/'.&declutter($m);
108: $m2=~s/\.meta$//;
109: my $q2="select * from metadata where url like binary '$m2'";
110: my $sth = $dbh->prepare($q2);
111: $sth->execute();
112: my $r1=$sth->fetchall_arrayref;
113: if (@$r1) {
114: $sth=$dbh->prepare("delete from metadata where url like binary '$m2'");
115: $sth->execute();
116: }
117: $sth=$dbh->prepare('insert into metadata values ('.
118: '"'.delete($ref->{'title'}).'"'.','.
119: '"'.delete($ref->{'author'}).'"'.','.
120: '"'.delete($ref->{'subject'}).'"'.','.
121: '"'.$m2.'"'.','.
122: '"'.delete($ref->{'keywords'}).'"'.','.
123: '"'.'current'.'"'.','.
124: '"'.delete($ref->{'notes'}).'"'.','.
125: '"'.delete($ref->{'abstract'}).'"'.','.
126: '"'.delete($ref->{'mime'}).'"'.','.
127: '"'.delete($ref->{'language'}).'"'.','.
128: '"'.sqltime(delete($ref->{'creationdate'})).'"'.','.
129: '"'.sqltime(delete($ref->{'lastrevisiondate'})).'"'.','.
130: '"'.delete($ref->{'owner'}).'"'.','.
131: '"'.delete($ref->{'copyright'}).'"'.')');
132: $sth->execute();
133: }
134:
135: # ----------------------------------------------------------- Clean up database
136: # Need to, perhaps, remove stale SQL database records.
137: # ... not yet implemented
138:
139: # --------------------------------------------------- Close database connection
140: $dbh->disconnect;
141:
142: # ---------------------------------------------------------------- Get metadata
143: # significantly altered from subroutine present in lonnet
144: sub metadata {
145: my ($uri,$what)=@_;
146: my %metacache;
147: $uri=&declutter($uri);
148: my $filename=$uri;
149: $uri=~s/\.meta$//;
150: $uri='';
151: unless ($metacache{$uri.'keys'}) {
152: unless ($filename=~/\.meta$/) { $filename.='.meta'; }
153: my $metastring=&getfile($perlvar{'lonDocRoot'}.'/res/'.$filename);
154: my $parser=HTML::TokeParser->new(\$metastring);
155: my $token;
156: while ($token=$parser->get_token) {
157: if ($token->[0] eq 'S') {
158: my $entry=$token->[1];
159: my $unikey=$entry;
160: if (defined($token->[2]->{'part'})) {
161: $unikey.='_'.$token->[2]->{'part'};
162: }
163: if (defined($token->[2]->{'name'})) {
164: $unikey.='_'.$token->[2]->{'name'};
165: }
166: if ($metacache{$uri.'keys'}) {
167: $metacache{$uri.'keys'}.=','.$unikey;
168: } else {
169: $metacache{$uri.'keys'}=$unikey;
170: }
171: map {
172: $metacache{$uri.''.$unikey.'.'.$_}=$token->[2]->{$_};
173: } @{$token->[3]};
174: unless (
175: $metacache{$uri.''.$unikey}=$parser->get_text('/'.$entry)
176: ) { $metacache{$uri.''.$unikey}=
177: $metacache{$uri.''.$unikey.'.default'};
178: }
179: }
180: }
181: }
182: return \%metacache;
183: }
184:
185: # ------------------------------------------------------------ Serves up a file
186: # returns either the contents of the file or a -1
187: sub getfile {
188: my $file=shift;
189: if (! -e $file ) { return -1; };
190: my $fh=IO::File->new($file);
191: my $a='';
192: while (<$fh>) { $a .=$_; }
193: return $a
194: }
195:
196: # ------------------------------------------------------------- Declutters URLs
197: sub declutter {
198: my $thisfn=shift;
199: $thisfn=~s/^$perlvar{'lonDocRoot'}//;
200: $thisfn=~s/^\///;
201: $thisfn=~s/^res\///;
202: return $thisfn;
203: }
204:
205: # --------------------------------------- Is this the home server of an author?
206: # (copied from lond, modification of the return value)
207: sub ishome {
208: my $author=shift;
209: $author=~s/\/home\/httpd\/html\/res\/([^\/]*)\/([^\/]*).*/$1\/$2/;
210: my ($udom,$uname)=split(/\//,$author);
211: my $proname=propath($udom,$uname);
212: if (-e $proname) {
213: return 1;
214: } else {
215: return 0;
216: }
217: }
218:
219: # -------------------------------------------- Return path to profile directory
220: # (copied from lond)
221: sub propath {
222: my ($udom,$uname)=@_;
223: $udom=~s/\W//g;
224: $uname=~s/\W//g;
225: my $subdir=$uname.'__';
226: $subdir =~ s/(.)(.)(.).*/$1\/$2\/$3/;
227: my $proname="$perlvar{'lonUsersDir'}/$udom/$subdir/$uname";
228: return $proname;
229: }
230:
231: # ---------------------------- convert 'time' format into a datetime sql format
232: sub sqltime {
233: my ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst) =
234: localtime(@_[0]);
235: $mon++; $year+=1900;
236: return "$year-$mon-$mday $hour:$min:$sec";
237: }
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>