Return to searchcat.pl CVS log

Up to [LON-CAPA] / loncom / metadata_database

Complete refactoring.

Currently does not attempt to get dynamic metadata.
Now uses lonmetadata routines to manage MySQL tables and inserts.
Appears to work fine, insert correct values into mysql, and run without
warnings or errors.
Added counting of copyright types.

#!/usr/bin/perl
# The LearningOnline Network
# searchcat.pl "Search Catalog" batch script
#
# $Id: searchcat.pl,v 1.55 2004/04/08 15:57:32 matthew Exp $
#
# Copyright Michigan State University Board of Trustees
#
# This file is part of the LearningOnline Network with CAPA (LON-CAPA).
#
# LON-CAPA is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# LON-CAPA is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with LON-CAPA; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
#
# /home/httpd/html/adm/gpl.txt
#
# http://www.lon-capa.org/
#
###

=pod

=head1 NAME

B<searchcat.pl> - put authoritative filesystem data into sql database.

=head1 SYNOPSIS

Ordinarily this script is to be called from a loncapa cron job
(CVS source location: F<loncapa/loncom/cron/loncapa>; typical
filesystem installation location: F</etc/cron.d/loncapa>).

Here is the cron job entry.

C<# Repopulate and refresh the metadata database used for the search catalog.>
C<10 1 * * 7    www    /home/httpd/perl/searchcat.pl>

This script only allows itself to be run as the user C<www>.

=head1 DESCRIPTION

This script goes through a loncapa resource directory and gathers metadata.
The metadata is entered into a SQL database.

This script also does general database maintenance such as reformatting
the C<loncapa:metadata> table if it is deprecated.

This script evaluates dynamic metadata from the authors'
F<nohist_resevaldata.db> database file in order to store it in MySQL.

This script is playing an increasingly important role for a loncapa
library server.  The proper operation of this script is critical for a smooth
and correct user experience.

=cut

use strict;

use DBI;
use lib '/home/httpd/lib/perl/';
use LONCAPA::Configuration;
use LONCAPA::lonmetadata;

use IO::File;
use HTML::TokeParser;
use GDBM_File;
use POSIX qw(strftime mktime);
use File::Find;

##
## Use variables for table names so we can test this routine a little easier
my $oldname = 'metadata';
my $newname = 'newmetadata';

#
# Read loncapa_apache.conf and loncapa.conf
my $perlvarref=LONCAPA::Configuration::read_conf('loncapa.conf');
my %perlvar=%{$perlvarref};
undef $perlvarref;
delete $perlvar{'lonReceipt'}; # remove since sensitive (really?) & not needed
#
# Only run if machine is a library server
exit if ($perlvar{'lonRole'} ne 'library');
#
#  Make sure this process is running from user=www
my $wwwid=getpwnam('www');
if ($wwwid!=$<) {
    my $emailto="$perlvar{'lonAdmEMail'},$perlvar{'lonSysEMail'}";
    my $subj="LON: $perlvar{'lonHostID'} User ID mismatch";
    system("echo 'User ID mismatch. searchcat.pl must be run as user www.' |\
 mailto $emailto -s '$subj' > /dev/null");
    exit 1;
}
#
# Let people know we are running
open(LOG,'>'.$perlvar{'lonDaemons'}.'/logs/searchcat.log');
print LOG '==== Searchcat Run '.localtime()."====\n";
#
# Connect to database
my $dbh;
if (! ($dbh = DBI->connect("DBI:mysql:loncapa","www",$perlvar{'lonSqlAccess'},
                          { RaiseError =>0,PrintError=>0}))) {
    print LOG "Cannot connect to database!\n";
    die "MySQL Error: Cannot connect to database!\n";
}
# This can return an error and still be okay, so we do not bother checking.
# (perhaps it should be more robust and check for specific errors)
$dbh->do('DROP TABLE IF EXISTS '.$newname);
#
# Create the new table
my $request = &LONCAPA::lonmetadata::create_metadata_storage($newname);
$dbh->do($request);
if ($dbh->err) {
    $dbh->disconnect();
    print LOG "\nMySQL Error Create: ".$dbh->errstr."\n";
    die $dbh->errstr;
}
#
# find out which users we need to examine
opendir(RESOURCES,"$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}");
my @homeusers = 
    grep {
        &ishome("$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}/$_");
    } grep { 
        !/^\.\.?$/;
    } readdir(RESOURCES);
closedir RESOURCES;
#
# Loop through the users
foreach my $user (@homeusers) {
    print LOG "=== User: ".$user."\n";
    my $prodir=&propath($perlvar{'lonDefDomain'},$user);
    #
    # Use File::Find to get the files we need to read/modify
    find(
         {preprocess => \&only_meta_files,
#          wanted     => \&print_filename,
#          wanted     => \&log_metadata,
          wanted     => \&process_meta_file,
          }, 
         "$perlvar{'lonDocRoot'}/res/$perlvar{'lonDefDomain'}/$user");
}
#
# Rename the table
$dbh->do('DROP TABLE IF EXISTS '.$oldname);
if (! $dbh->do('RENAME TABLE '.$newname.' TO '.$oldname)) {
    print LOG "MySQL Error Rename: ".$dbh->errstr."\n";
    die $dbh->errstr;
}
if (! $dbh->disconnect) {
    print LOG "MySQL Error Disconnect: ".$dbh->errstr."\n";
    die $dbh->errstr;
}
##
## Finished!
print LOG "==== Searchcat completed ".localtime()." ====\n";
close(LOG);

&write_type_count();
&write_copyright_count();

exit 0;

########################################################
########################################################
###                                                  ###
###          File::Find support routines             ###
###                                                  ###
########################################################
########################################################
##
## &only_meta_files
##
## Called by File::Find.
## Takes a list of files/directories in and returns a list of files/directories
## to search.
sub only_meta_files {
    my @PossibleFiles = @_;
    my @ChosenFiles;
    foreach my $file (@PossibleFiles) {
        if ( ($file =~ /\.meta$/ &&            # Ends in meta
              $file !~ /\.\d+\.[^\.]+\.meta$/  # is not for a prior version
             ) || (-d $file )) { # directories are okay
                 # but we do not want /. or /..
            push(@ChosenFiles,$file);
        }
    }
    return @ChosenFiles;
}

##
##
## Debugging routines, use these for 'wanted' in the File::Find call
##
sub print_filename {
    my ($file) = $_;
    my $fullfilename = $File::Find::name;
    if (-d $file) {
        print LOG " Got directory ".$fullfilename."\n";
    } else {
        print LOG " Got file ".$fullfilename."\n";
    }
    $_=$file;
}

sub log_metadata {
    my ($file) = $_;
    my $fullfilename = $File::Find::name;
    return if (-d $fullfilename); # No need to do anything here for directories
    print LOG $fullfilename."\n";
    my $ref=&metadata($fullfilename);
    if (! defined($ref)) {
        print LOG "    No data\n";
        return;
    }
    while (my($key,$value) = each(%$ref)) {
        print LOG "    ".$key." => ".$value."\n";
    }
    &count_copyright($ref->{'copyright'});
    $_=$file;
}


##
## process_meta_file
##   Called by File::Find. 
##   Only input is the filename in $_.  
sub process_meta_file {
    my ($file) = $_;
    my $filename = $File::Find::name;
    return if (-d $filename); # No need to do anything here for directories
    #
    print LOG $filename."\n";
    #
    my $ref=&metadata($filename);
    #
    # $url is the original file url, not the metadata file
    my $url='/res/'.&declutter($filename);
    $url=~s/\.meta$//;
    print LOG "    ".$url."\n";
    #
    # Ignore some files based on their metadata
    if ($ref->{'obsolete'}) { 
        print LOG "obsolete\n"; 
        return; 
    }
    &count_copyright($ref->{'copyright'});
    if ($ref->{'copyright'} eq 'private') { 
        print LOG "private\n"; 
        return; 
    }
    #
    # Find the dynamic metadata
    my %dyn;
    if ($url=~ m:/default$:) {
        $url=~ s:/default$:/:;
    } else {
        # %dyn=&dynamicmeta($url);
        &count_type($url);
    }
    #
    $ref->{'creationdate'}     = &sqltime($ref->{'creationdate'});
    $ref->{'lastrevisiondate'} = &sqltime($ref->{'lastrevisiondate'});
    my %Data = (
                %$ref,
                %dyn,
                'url'=>$url,
                'version'=>'current');
    my ($count,$err) = &LONCAPA::lonmetadata::store_metadata($dbh,$newname,
                                                             \%Data);
    if ($err) {
        print LOG "\nMySQL Error Insert: ".$err."\n";
        die $err;
    }
    if ($count < 1) {
        print LOG "Unable to insert record into MySQL database for $url\n";
        die "Unable to insert record into MySQl database for $url";
    } else {
        print LOG "Count = ".$count."\n";
    }
    #
    # Reset $_ before leaving
    $_ = $file;
}

########################################################
########################################################
###                                                  ###
###  &metadata($uri)                                 ###
###   Retrieve metadata for the given file           ###
###                                                  ###
########################################################
########################################################
sub metadata {
    my ($uri)=@_;
    my %metacache=();
    $uri=&declutter($uri);
    my $filename=$uri;
    $uri=~s/\.meta$//;
    $uri='';
    if ($filename !~ /\.meta$/) { 
        $filename.='.meta';
    }
    my $metastring=&getfile($perlvar{'lonDocRoot'}.'/res/'.$filename);
    return undef if (! defined($metastring));
    my $parser=HTML::TokeParser->new(\$metastring);
    my $token;
    while ($token=$parser->get_token) {
        if ($token->[0] eq 'S') {
            my $entry=$token->[1];
            my $unikey=$entry;
            if (defined($token->[2]->{'part'})) { 
                $unikey.='_'.$token->[2]->{'part'}; 
            }
            if (defined($token->[2]->{'name'})) { 
                $unikey.='_'.$token->[2]->{'name'}; 
            }
            if ($metacache{$uri.'keys'}) {
                $metacache{$uri.'keys'}.=','.$unikey;
            } else {
                $metacache{$uri.'keys'}=$unikey;
            }
            foreach ( @{$token->[3]}) {
                $metacache{$uri.''.$unikey.'.'.$_}=$token->[2]->{$_};
            } 
            if (! ($metacache{$uri.''.$unikey}=$parser->get_text('/'.$entry))){
                $metacache{$uri.''.$unikey} = 
                    $metacache{$uri.''.$unikey.'.default'};
            }
        } # End of ($token->[0] eq 'S')
    }
    return \%metacache;
}

##
## &getfile($filename)
##   Slurps up an entire file into a scalar.  
##   Returns undef if the file does not exist
sub getfile {
    my $file = shift();
    if (! -e $file ) { 
        return undef; 
    }
    my $fh=IO::File->new($file);
    my $contents = '';
    while (<$fh>) { 
        $contents .= $_;
    }
    return $contents;
}

########################################################
########################################################
###                                                  ###
###    Dynamic Metadata                              ###
###                                                  ###
########################################################
########################################################
sub dynamicmeta {
    my $url = &declutter(shift());
    $url =~ s/\.meta$//;
    my %data = ('count'         => 0,
                'course'        => 0,
                'course_list'   => '',
                'avetries'      => 'NULL',
                'avetries_list' => '',
                'stdno'         => 0,
                'stdno_list'    => '',
                'usage'         => 0,
                'usage_list'    => '',
                'goto'          => 0,
                'goto_list'     => '',
                'comefrom'      => 0,
                'comefrom_list' => '',
                'difficulty'    => 'NULL',
                'difficulty_list' => '',
                'sequsage'      => '0',
                'sequsage_list' => '',
                'clear'         => 'NULL',
                'technical'     => 'NULL',
                'correct'       => 'NULL',
                'helpful'       => 'NULL',
                'depth'         => 'NULL',
                'comments'      => '',                
                );
    my ($dom,$auth)=($url=~/^(\w+)\/(\w+)\//);
    my $prodir=&propath($dom,$auth);
    #
    # Get metadata except counts
    my %evaldata;
    if (! tie(%evaldata,'GDBM_File',
              $prodir.'/nohist_resevaldata.db',&GDBM_READER(),0640)) {
        return (undef);
    }
    my %sum=();
    my %count=();
    my %concat=();
    my %listitems=(
                   'course'       => 'add',
                   'goto'         => 'add',
                   'comefrom'     => 'add',
                   'avetries'     => 'average',
                   'stdno'        => 'add',
                   'difficulty'   => 'average',
                   'clear'        => 'average',
                   'technical'    => 'average',
                   'helpful'      => 'average',
                   'correct'      => 'average',
                   'depth'        => 'average',
                   'comments'     => 'append',
                   'usage'        => 'count'
                   );
    #
    my $regexp=$url;
    $regexp=~s/(\W)/\\$1/g;
    $regexp='___'.$regexp.'___([a-z]+)$';
    while (my ($esckey,$value)=each %evaldata) {
        my $key=&unescape($esckey);
        if ($key=~/$regexp/) {
            my ($item,$purl,$cat)=split(/___/,$key);
            $count{$cat}++;
            if ($listitems{$cat} ne 'append') {
                if (defined($sum{$cat})) {
                    $sum{$cat}+=&unescape($value);
                    $concat{$cat}.=','.$item;
                } else {
                    $sum{$cat}=&unescape($value);
                    $concat{$cat}=$item;
                }
            } else {
                if (defined($sum{$cat})) {
                    if ($evaldata{$esckey}=~/\w/) {
                        $sum{$cat}.='<hr />'.&unescape($evaldata{$esckey});
                    }
                } else {
                    $sum{$cat}=''.&unescape($evaldata{$esckey});
		    }
            }
        }
    }
    untie(%evaldata);
    # transfer gathered data to returnhash, calculate averages where applicable
    my %returnhash;
    while (my $cat=each(%count)) {
        if ($count{$cat} eq 'nan') { next; }
        if ($sum{$cat} eq 'nan') { next; }
        if ($listitems{$cat} eq 'average') {
            if ($count{$cat}) {
                $returnhash{$cat}=int(($sum{$cat}/$count{$cat})*100.0+0.5)/100.0;
            } else {
                $returnhash{$cat}='NULL';
            }
        } elsif ($listitems{$cat} eq 'count') {
            $returnhash{$cat}=$count{$cat};
        } else {
            $returnhash{$cat}=$sum{$cat};
        }
        $returnhash{$cat.'_list'}=$concat{$cat};
    }
    #
    # get count
    if (tie(my %evaldata,'GDBM_File',
            $prodir.'/nohist_accesscount.db',&GDBM_READER(),0640)) {
	my $escurl=&escape($url);
	if (! exists($evaldata{$escurl})) {
	    $returnhash{'count'}=0;
	} else {
	    $returnhash{'count'}=$evaldata{$escurl};
	}
	untie %evaldata;
    }
    return %returnhash;
}

########################################################
########################################################
###                                                  ###
###   Counts                                         ###
###                                                  ###
########################################################
########################################################
{

my %countext;

sub count_type {
    my $file=shift;
    $file=~/\.(\w+)$/;
    my $ext=lc($1);
    $countext{$ext}++;
}

sub write_type_count {
    open(RESCOUNT,'>/home/httpd/html/lon-status/rescount.txt');
    while (my ($extension,$count) = each(%countext)) {
	print RESCOUNT $extension.'='.$count.'&';
    }
    print RESCOUNT 'time='.time."\n";
    close(RESCOUNT);
}

} # end of scope for %countext

{

my %copyrights;

sub count_copyright {
    $copyrights{@_[0]}++;
}

sub write_copyright_count {
    open(COPYCOUNT,'>/home/httpd/html/lon-status/copyrightcount.txt');
    while (my ($copyright,$count) = each(%copyrights)) {
	print COPYCOUNT $copyright.'='.$count.'&';
    }
    print COPYCOUNT 'time='.time."\n";
    close(COPYCOUNT);
}

} # end of scope for %copyrights

########################################################
########################################################
###                                                  ###
###   Miscellanous Utility Routines                  ###
###                                                  ###
########################################################
########################################################
##
## &ishome($username)
##   Returns 1 if $username is a LON-CAPA author, 0 otherwise
##   (copied from lond, modification of the return value)
sub ishome {
    my $author=shift;
    $author=~s/\/home\/httpd\/html\/res\/([^\/]*)\/([^\/]*).*/$1\/$2/;
    my ($udom,$uname)=split(/\//,$author);
    my $proname=propath($udom,$uname);
    if (-e $proname) {
	return 1;
    } else {
        return 0;
    }
}

##
## &propath($udom,$uname)
##   Returns the path to the users LON-CAPA directory
##   (copied from lond)
sub propath {
    my ($udom,$uname)=@_;
    $udom=~s/\W//g;
    $uname=~s/\W//g;
    my $subdir=$uname.'__';
    $subdir =~ s/(.)(.)(.).*/$1\/$2\/$3/;
    my $proname="$perlvar{'lonUsersDir'}/$udom/$subdir/$uname";
    return $proname;
} 

##
## &sqltime($timestamp)
##
## Convert perl $timestamp to MySQL time.  MySQL expects YYYY-MM-DD HH:MM:SS
##
sub sqltime {
    my ($time) = @_;
    my $mysqltime;
    if ($time =~ 
        /(\d+)-(\d+)-(\d+) # YYYY-MM-DD
        \s                 # a space
        (\d+):(\d+):(\d+)  # HH:MM::SS
        /x ) { 
        # Some of the .meta files have the time in mysql
        # format already, so just make sure they are 0 padded and
        # pass them back.
        $mysqltime = sprintf('%04d-%02d-%02d %02d:%02d:%02d',
                             $1,$2,$3,$4,$5,$6);
    } elsif ($time =~ /^\d+$/) {
        my @TimeData = gmtime($time);
        # Alter the month to be 1-12 instead of 0-11
        $TimeData[4]++;
        # Alter the year to be from 0 instead of from 1900
        $TimeData[5]+=1900;
        $mysqltime = sprintf('%04d-%02d-%02d %02d:%02d:%02d',
                             @TimeData[5,4,3,2,1,0]);
    } else {
        print LOG "    Unable to decode time ".$time."\n";
        $mysqltime = 0;
    }
    return $mysqltime;
}

##
## &declutter($filename)
##   Given a filename, returns a url for the filename.
sub declutter {
    my $thisfn=shift;
    $thisfn=~s/^$perlvar{'lonDocRoot'}//;
    $thisfn=~s/^\///;
    $thisfn=~s/^res\///;
    return $thisfn;
}

##
## Escape / Unescape special characters
sub unescape {
    my $str=shift;
    $str =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C",hex($1))/eg;
    return $str;
}

sub escape {
    my $str=shift;
    $str =~ s/(\W)/"%".unpack('H2',$1)/eg;
    return $str;
}