Annotation of loncom/metadata_database/searchcat.pl, revision 1.22

1.1       harris41    1: #!/usr/bin/perl
                      2: # The LearningOnline Network
                      3: # searchcat.pl "Search Catalog" batch script
1.16      harris41    4: #
1.22    ! www         5: # $Id: searchcat.pl,v 1.21 2002/10/08 15:09:36 www Exp $
1.16      harris41    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
1.17      harris41   30: # 04/14/2001, 04/16/2001 Scott Harrison
1.16      harris41   31: #
1.17      harris41   32: # YEAR=2002
                     33: # 05/11/2002 Scott Harrison
1.16      harris41   34: #
                     35: ###
1.1       harris41   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: 
1.17      harris41   41: use lib '/home/httpd/lib/perl/';
                     42: use LONCAPA::Configuration;
                     43: 
1.1       harris41   44: use IO::File;
                     45: use HTML::TokeParser;
1.6       harris41   46: use DBI;
1.21      www        47: use GDBM_File;
1.1       harris41   48: 
                     49: my @metalist;
1.21      www        50: 
                     51: 
                     52: # ----------------------------------------------------- Un-Escape Special Chars
                     53: 
                     54: sub unescape {
                     55:     my $str=shift;
                     56:     $str =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C",hex($1))/eg;
                     57:     return $str;
                     58: }
                     59: 
1.22    ! www        60: # -------------------------------------------------------- Escape Special Chars
        !            61: 
        !            62: sub escape {
        !            63:     my $str=shift;
        !            64:     $str =~ s/(\W)/"%".unpack('H2',$1)/eg;
        !            65:     return $str;
        !            66: }
        !            67: 
1.21      www        68: 
                     69: # ------------------------------------------- Code to evaluate dynamic metadata
                     70: 
                     71: sub dynamicmeta {
                     72: #
                     73: #
                     74: # Do nothing for now ...
                     75: #
                     76: #
                     77:     return;
                     78: #
                     79: # ..., but stuff below already works
                     80: #
                     81:     my $url=&declutter(shift);
                     82:     $url=~s/\.meta$//;
                     83:     my %returnhash=();
                     84:     my ($adomain,$aauthor)=($url=~/^(\w+)\/(\w+)\//);
                     85:     my $prodir=&propath($adomain,$aauthor);
                     86:     if (tie(%evaldata,'GDBM_File',
1.22    ! www        87:             $prodir.'/nohist_resevaldata.db',&GDBM_WRCREAT,0640)) {
1.21      www        88:        my %sum=();
                     89:        my %cnt=();
                     90:        my %listitems=('count'        => 'add',
                     91:                       'course'       => 'add',
                     92:                       'avetries'     => 'avg',
                     93:                       'stdno'        => 'add',
                     94:                       'difficulty'   => 'avg',
                     95:                       'clear'        => 'avg',
                     96:                       'technical'    => 'avg',
                     97:                       'helpful'      => 'avg',
                     98:                       'correct'      => 'avg',
                     99:                       'depth'        => 'avg',
                    100:                       'comments'     => 'app',
                    101:                       'usage'        => 'cnt'
                    102:                       );
                    103:        my $regexp=$url;
                    104:        $regexp=~s/(\W)/\\$1/g;
                    105:        $regexp='___'.$regexp.'___([a-z]+)$';
                    106:        foreach (keys %evaldata) {
                    107: 	 my $key=&unescape($_);
                    108: 	 if ($key=~/$regexp/) {
1.22    ! www       109: 	    my $ctype=$1;
        !           110:             if (defined($cnt{$ctype})) { 
        !           111:                $cnt{$ctype}++; 
        !           112:             } else { 
        !           113:                $cnt{$ctype}=1; 
        !           114:             }
        !           115:             unless ($listitems{$ctype} eq 'app') {
        !           116:                if (defined($sum{$ctype})) {
        !           117:                   $sum{$ctype}+=$evaldata{$_};
        !           118:    	       } else {
        !           119:                   $sum{$ctype}=$evaldata{$_};
        !           120: 	       }
        !           121:             } else {
        !           122:                if (defined($sum{$ctype})) {
        !           123:                   if ($evaldata{$_}) {
        !           124:                      $sum{$ctype}.='<hr>'.$evaldata{$_};
1.21      www       125: 	          }
1.22    ! www       126:  	       } else {
        !           127: 	             $sum{$ctype}=''.$evaldata{$_};
        !           128: 	       }
        !           129: 	    }
        !           130: 	    if ($ctype eq 'count') {
        !           131: 	       delete($evaldata{$_});
        !           132:             }
        !           133: 	 }
        !           134:       }
        !           135:       foreach (keys %cnt) {
        !           136:          if ($listitems{$_} eq 'avg') {
        !           137: 	     $returnhash{$_}=int(($sum{$_}/$cnt{$_})*100.0+0.5)/100.0;
        !           138:          } elsif ($listitems{$_} eq 'cnt') {
        !           139:              $returnhash{$_}=$cnt{$_};
        !           140:          } else {
        !           141:              $returnhash{$_}=$sum{$_};
        !           142:          }
        !           143:      }
        !           144:      if ($returnhash{'count'}) {
        !           145:          my $newkey=$$.'_'.time.'_searchcat___'.&escape($url).'___count';
        !           146:          $evaldata{$newkey}=$returnhash{'count'};
1.21      www       147:      }
                    148:      untie(%evaldata);
                    149:    }
                    150:    return %returnhash;
                    151: }
                    152:   
1.1       harris41  153: # ----------------- Code to enable 'find' subroutine listing of the .meta files
                    154: require "find.pl";
                    155: sub wanted {
                    156:     (($dev,$ino,$mode,$nlink,$uid,$gid) = lstat($_)) &&
                    157:     -f _ &&
1.10      harris41  158:     /^.*\.meta$/ && !/^.+\.\d+\.[^\.]+\.meta$/ &&
1.1       harris41  159:     push(@metalist,"$dir/$_");
                    160: }
                    161: 
1.18      matthew   162: # ---------------  Read loncapa_apache.conf and loncapa.conf and get variables
1.20      harris41  163: my $perlvarref=LONCAPA::Configuration::read_conf('loncapa.conf');
1.17      harris41  164: my %perlvar=%{$perlvarref};
                    165: undef $perlvarref; # remove since sensitive and not needed
                    166: delete $perlvar{'lonReceipt'}; # remove since sensitive and not needed
1.15      harris41  167: 
                    168: # ------------------------------------- Only run if machine is a library server
                    169: exit unless $perlvar{'lonRole'} eq 'library';
1.1       harris41  170: 
1.3       harris41  171: my $dbh;
1.1       harris41  172: # ------------------------------------- Make sure that database can be accessed
                    173: {
                    174:     unless (
                    175: 	    $dbh = DBI->connect("DBI:mysql:loncapa","www",$perlvar{'lonSqlAccess'},{ RaiseError =>0,PrintError=>0})
                    176: 	    ) { 
                    177: 	print "Cannot connect to database!\n";
                    178: 	exit;
                    179:     }
1.19      matthew   180:     my $make_metadata_table = "CREATE TABLE IF NOT EXISTS metadata (".
                    181:         "title TEXT, author TEXT, subject TEXT, url TEXT, keywords TEXT, ".
                    182:         "version TEXT, notes TEXT, abstract TEXT, mime TEXT, language TEXT, ".
                    183:         "creationdate DATETIME, lastrevisiondate DATETIME, owner TEXT, ".
                    184:         "copyright TEXT, FULLTEXT idx_title (title), ".
                    185:         "FULLTEXT idx_author (author), FULLTEXT idx_subject (subject), ".
                    186:         "FULLTEXT idx_url (url), FULLTEXT idx_keywords (keywords), ".
                    187:         "FULLTEXT idx_version (version), FULLTEXT idx_notes (notes), ".
                    188:         "FULLTEXT idx_abstract (abstract), FULLTEXT idx_mime (mime), ".
                    189:         "FULLTEXT idx_language (language), FULLTEXT idx_owner (owner), ".
                    190:         "FULLTEXT idx_copyright (copyright)) TYPE=MYISAM";
                    191:     # It would sure be nice to have some logging mechanism.
                    192:     $dbh->do($make_metadata_table);
1.1       harris41  193: }
                    194: 
                    195: # ------------------------------------------------------------- get .meta files
1.2       harris41  196: opendir(RESOURCES,"$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}");
                    197: my @homeusers=grep
                    198:           {&ishome("$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}/$_")}
                    199:           grep {!/^\.\.?$/} readdir(RESOURCES);
                    200: closedir RESOURCES;
                    201: foreach my $user (@homeusers) {
                    202:     &find("$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}/$user");
                    203: }
1.1       harris41  204: 
                    205: # -- process each file to get metadata and put into search catalog SQL database
1.9       harris41  206: # Also, check to see if already there.
1.11      harris41  207: # I could just delete (without searching first), but this works for now.
1.1       harris41  208: foreach my $m (@metalist) {
                    209:     my $ref=&metadata($m);
1.11      harris41  210:     my $m2='/res/'.&declutter($m);
1.12      harris41  211:     $m2=~s/\.meta$//;
1.21      www       212:     &dynamicmeta($m2);
1.12      harris41  213:     my $q2="select * from metadata where url like binary '$m2'";
1.9       harris41  214:     my $sth = $dbh->prepare($q2);
                    215:     $sth->execute();
                    216:     my $r1=$sth->fetchall_arrayref;
                    217:     if (@$r1) {
1.12      harris41  218: 	$sth=$dbh->prepare("delete from metadata where url like binary '$m2'");
1.9       harris41  219:         $sth->execute();
                    220:     }
                    221:     $sth=$dbh->prepare('insert into metadata values ('.
1.8       harris41  222: 			  '"'.delete($ref->{'title'}).'"'.','.
                    223: 			  '"'.delete($ref->{'author'}).'"'.','.
                    224: 			  '"'.delete($ref->{'subject'}).'"'.','.
1.12      harris41  225: 			  '"'.$m2.'"'.','.
1.8       harris41  226: 			  '"'.delete($ref->{'keywords'}).'"'.','.
1.9       harris41  227: 			  '"'.'current'.'"'.','.
1.8       harris41  228: 			  '"'.delete($ref->{'notes'}).'"'.','.
                    229: 			  '"'.delete($ref->{'abstract'}).'"'.','.
                    230: 			  '"'.delete($ref->{'mime'}).'"'.','.
                    231: 			  '"'.delete($ref->{'language'}).'"'.','.
1.13      harris41  232: 			  '"'.sqltime(delete($ref->{'creationdate'})).'"'.','.
                    233: 			  '"'.sqltime(delete($ref->{'lastrevisiondate'})).'"'.','.
1.8       harris41  234: 			  '"'.delete($ref->{'owner'}).'"'.','.
                    235: 			  '"'.delete($ref->{'copyright'}).'"'.')');
1.1       harris41  236:     $sth->execute();
                    237: }
                    238: 
                    239: # ----------------------------------------------------------- Clean up database
                    240: # Need to, perhaps, remove stale SQL database records.
                    241: # ... not yet implemented
                    242: 
                    243: # --------------------------------------------------- Close database connection
                    244: $dbh->disconnect;
                    245: 
                    246: # ---------------------------------------------------------------- Get metadata
                    247: # significantly altered from subroutine present in lonnet
                    248: sub metadata {
                    249:     my ($uri,$what)=@_;
                    250:     my %metacache;
                    251:     $uri=&declutter($uri);
                    252:     my $filename=$uri;
                    253:     $uri=~s/\.meta$//;
                    254:     $uri='';
                    255:     unless ($metacache{$uri.'keys'}) {
                    256:         unless ($filename=~/\.meta$/) { $filename.='.meta'; }
                    257: 	my $metastring=&getfile($perlvar{'lonDocRoot'}.'/res/'.$filename);
                    258:         my $parser=HTML::TokeParser->new(\$metastring);
                    259:         my $token;
                    260:         while ($token=$parser->get_token) {
                    261:            if ($token->[0] eq 'S') {
                    262: 	      my $entry=$token->[1];
                    263:               my $unikey=$entry;
                    264:               if (defined($token->[2]->{'part'})) { 
                    265:                  $unikey.='_'.$token->[2]->{'part'}; 
                    266: 	      }
                    267:               if (defined($token->[2]->{'name'})) { 
                    268:                  $unikey.='_'.$token->[2]->{'name'}; 
                    269: 	      }
                    270:               if ($metacache{$uri.'keys'}) {
                    271:                  $metacache{$uri.'keys'}.=','.$unikey;
                    272:               } else {
                    273:                  $metacache{$uri.'keys'}=$unikey;
                    274: 	      }
                    275:               map {
                    276: 		  $metacache{$uri.''.$unikey.'.'.$_}=$token->[2]->{$_};
                    277:               } @{$token->[3]};
                    278:               unless (
                    279:                  $metacache{$uri.''.$unikey}=$parser->get_text('/'.$entry)
                    280: 		      ) { $metacache{$uri.''.$unikey}=
                    281: 			      $metacache{$uri.''.$unikey.'.default'};
                    282: 		      }
                    283:           }
                    284:        }
                    285:     }
                    286:     return \%metacache;
                    287: }
                    288: 
                    289: # ------------------------------------------------------------ Serves up a file
                    290: # returns either the contents of the file or a -1
                    291: sub getfile {
                    292:   my $file=shift;
                    293:   if (! -e $file ) { return -1; };
                    294:   my $fh=IO::File->new($file);
                    295:   my $a='';
                    296:   while (<$fh>) { $a .=$_; }
                    297:   return $a
                    298: }
                    299: 
                    300: # ------------------------------------------------------------- Declutters URLs
                    301: sub declutter {
                    302:     my $thisfn=shift;
                    303:     $thisfn=~s/^$perlvar{'lonDocRoot'}//;
                    304:     $thisfn=~s/^\///;
                    305:     $thisfn=~s/^res\///;
                    306:     return $thisfn;
                    307: }
1.2       harris41  308: 
                    309: # --------------------------------------- Is this the home server of an author?
                    310: # (copied from lond, modification of the return value)
                    311: sub ishome {
                    312:     my $author=shift;
                    313:     $author=~s/\/home\/httpd\/html\/res\/([^\/]*)\/([^\/]*).*/$1\/$2/;
                    314:     my ($udom,$uname)=split(/\//,$author);
                    315:     my $proname=propath($udom,$uname);
                    316:     if (-e $proname) {
                    317: 	return 1;
                    318:     } else {
                    319:         return 0;
                    320:     }
                    321: }
                    322: 
                    323: # -------------------------------------------- Return path to profile directory
                    324: # (copied from lond)
                    325: sub propath {
                    326:     my ($udom,$uname)=@_;
                    327:     $udom=~s/\W//g;
                    328:     $uname=~s/\W//g;
                    329:     my $subdir=$uname.'__';
                    330:     $subdir =~ s/(.)(.)(.).*/$1\/$2\/$3/;
                    331:     my $proname="$perlvar{'lonUsersDir'}/$udom/$subdir/$uname";
                    332:     return $proname;
                    333: } 
1.13      harris41  334: 
                    335: # ---------------------------- convert 'time' format into a datetime sql format
                    336: sub sqltime {
                    337:     my ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst) =
                    338: 	localtime(@_[0]);
1.14      harris41  339:     $mon++; $year+=1900;
1.13      harris41  340:     return "$year-$mon-$mday $hour:$min:$sec";
                    341: }

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>