1: #!/usr/bin/perl
2: # The LearningOnline Network
3: # searchcat.pl "Search Catalog" batch script
4: #
5: # $Id: searchcat.pl,v 1.17 2002/05/11 21:28:20 harris41 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 access.conf and loncapa.conf and get variables
59: my $perlvarref=LONCAPA::Configuration::read_conf('access.conf','loncapa.conf');
60: my %perlvar=%{$perlvarref};
61: undef $perlvarref; # remove since sensitive and not needed
62: delete $perlvar{'lonReceipt'}; # remove since sensitive and not needed
63:
64: # ------------------------------------- Only run if machine is a library server
65: exit unless $perlvar{'lonRole'} eq 'library';
66:
67: my $dbh;
68: # ------------------------------------- Make sure that database can be accessed
69: {
70: unless (
71: $dbh = DBI->connect("DBI:mysql:loncapa","www",$perlvar{'lonSqlAccess'},{ RaiseError =>0,PrintError=>0})
72: ) {
73: print "Cannot connect to database!\n";
74: exit;
75: }
76: }
77:
78: # ------------------------------------------------------------- get .meta files
79: opendir(RESOURCES,"$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}");
80: my @homeusers=grep
81: {&ishome("$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}/$_")}
82: grep {!/^\.\.?$/} readdir(RESOURCES);
83: closedir RESOURCES;
84: foreach my $user (@homeusers) {
85: &find("$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}/$user");
86: }
87:
88: # -- process each file to get metadata and put into search catalog SQL database
89: # Also, check to see if already there.
90: # I could just delete (without searching first), but this works for now.
91: foreach my $m (@metalist) {
92: my $ref=&metadata($m);
93: my $m2='/res/'.&declutter($m);
94: $m2=~s/\.meta$//;
95: my $q2="select * from metadata where url like binary '$m2'";
96: my $sth = $dbh->prepare($q2);
97: $sth->execute();
98: my $r1=$sth->fetchall_arrayref;
99: if (@$r1) {
100: $sth=$dbh->prepare("delete from metadata where url like binary '$m2'");
101: $sth->execute();
102: }
103: $sth=$dbh->prepare('insert into metadata values ('.
104: '"'.delete($ref->{'title'}).'"'.','.
105: '"'.delete($ref->{'author'}).'"'.','.
106: '"'.delete($ref->{'subject'}).'"'.','.
107: '"'.$m2.'"'.','.
108: '"'.delete($ref->{'keywords'}).'"'.','.
109: '"'.'current'.'"'.','.
110: '"'.delete($ref->{'notes'}).'"'.','.
111: '"'.delete($ref->{'abstract'}).'"'.','.
112: '"'.delete($ref->{'mime'}).'"'.','.
113: '"'.delete($ref->{'language'}).'"'.','.
114: '"'.sqltime(delete($ref->{'creationdate'})).'"'.','.
115: '"'.sqltime(delete($ref->{'lastrevisiondate'})).'"'.','.
116: '"'.delete($ref->{'owner'}).'"'.','.
117: '"'.delete($ref->{'copyright'}).'"'.')');
118: $sth->execute();
119: }
120:
121: # ----------------------------------------------------------- Clean up database
122: # Need to, perhaps, remove stale SQL database records.
123: # ... not yet implemented
124:
125: # --------------------------------------------------- Close database connection
126: $dbh->disconnect;
127:
128: # ---------------------------------------------------------------- Get metadata
129: # significantly altered from subroutine present in lonnet
130: sub metadata {
131: my ($uri,$what)=@_;
132: my %metacache;
133: $uri=&declutter($uri);
134: my $filename=$uri;
135: $uri=~s/\.meta$//;
136: $uri='';
137: unless ($metacache{$uri.'keys'}) {
138: unless ($filename=~/\.meta$/) { $filename.='.meta'; }
139: my $metastring=&getfile($perlvar{'lonDocRoot'}.'/res/'.$filename);
140: my $parser=HTML::TokeParser->new(\$metastring);
141: my $token;
142: while ($token=$parser->get_token) {
143: if ($token->[0] eq 'S') {
144: my $entry=$token->[1];
145: my $unikey=$entry;
146: if (defined($token->[2]->{'part'})) {
147: $unikey.='_'.$token->[2]->{'part'};
148: }
149: if (defined($token->[2]->{'name'})) {
150: $unikey.='_'.$token->[2]->{'name'};
151: }
152: if ($metacache{$uri.'keys'}) {
153: $metacache{$uri.'keys'}.=','.$unikey;
154: } else {
155: $metacache{$uri.'keys'}=$unikey;
156: }
157: map {
158: $metacache{$uri.''.$unikey.'.'.$_}=$token->[2]->{$_};
159: } @{$token->[3]};
160: unless (
161: $metacache{$uri.''.$unikey}=$parser->get_text('/'.$entry)
162: ) { $metacache{$uri.''.$unikey}=
163: $metacache{$uri.''.$unikey.'.default'};
164: }
165: }
166: }
167: }
168: return \%metacache;
169: }
170:
171: # ------------------------------------------------------------ Serves up a file
172: # returns either the contents of the file or a -1
173: sub getfile {
174: my $file=shift;
175: if (! -e $file ) { return -1; };
176: my $fh=IO::File->new($file);
177: my $a='';
178: while (<$fh>) { $a .=$_; }
179: return $a
180: }
181:
182: # ------------------------------------------------------------- Declutters URLs
183: sub declutter {
184: my $thisfn=shift;
185: $thisfn=~s/^$perlvar{'lonDocRoot'}//;
186: $thisfn=~s/^\///;
187: $thisfn=~s/^res\///;
188: return $thisfn;
189: }
190:
191: # --------------------------------------- Is this the home server of an author?
192: # (copied from lond, modification of the return value)
193: sub ishome {
194: my $author=shift;
195: $author=~s/\/home\/httpd\/html\/res\/([^\/]*)\/([^\/]*).*/$1\/$2/;
196: my ($udom,$uname)=split(/\//,$author);
197: my $proname=propath($udom,$uname);
198: if (-e $proname) {
199: return 1;
200: } else {
201: return 0;
202: }
203: }
204:
205: # -------------------------------------------- Return path to profile directory
206: # (copied from lond)
207: sub propath {
208: my ($udom,$uname)=@_;
209: $udom=~s/\W//g;
210: $uname=~s/\W//g;
211: my $subdir=$uname.'__';
212: $subdir =~ s/(.)(.)(.).*/$1\/$2\/$3/;
213: my $proname="$perlvar{'lonUsersDir'}/$udom/$subdir/$uname";
214: return $proname;
215: }
216:
217: # ---------------------------- convert 'time' format into a datetime sql format
218: sub sqltime {
219: my ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst) =
220: localtime(@_[0]);
221: $mon++; $year+=1900;
222: return "$year-$mon-$mday $hour:$min:$sec";
223: }
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>