# The LearningOnline Network with CAPA # (Publication Handler # # $Id: loncoursedata.pm,v 1.29 2002/09/03 02:22:40 albertel 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 loncoursedata =head1 SYNOPSIS Set of functions that download and process student and course information. =head1 PACKAGES USED Apache::Constants qw(:common :http) Apache::lonnet() Apache::lonhtmlcommon HTML::TokeParser GDBM_File =cut package Apache::loncoursedata; use strict; use Apache::Constants qw(:common :http); use Apache::lonnet(); use Apache::lonhtmlcommon; use HTML::TokeParser; use GDBM_File; =pod =head1 DOWNLOAD INFORMATION This section contains all the functions that get data from other servers and/or itself. =cut # ----- DOWNLOAD INFORMATION ------------------------------------------- =pod =item &DownloadClasslist() Collects lastname, generation, middlename, firstname, PID, and section for each student from their environment database. The section data is also download, though it is in a rough format, and is processed later. The list of students is built from collecting a classlist for the course that is to be displayed. Once the classlist has been downloaded, its date stamp is recorded. Unless the datestamp for the class database is reset or is modified, this data will not be downloaded again. Also, there was talk about putting the fullname and section and perhaps other pieces of data into the classlist file. This would reduce the number of different file accesses and reduce the amount of processing on this side. =over 4 Input: $courseID, $lastDownloadTime, $c $courseID: The id of the course $lastDownloadTime: This is the date stamp for when this information was last gathered. If it is set to Not downloaded, it will gather the data again, though it currently does not remove the old data. $c: The connection class that can determine if the browser has aborted. It is used to short circuit this function so that it does not continue to get information when there is no need. Output: \%classlist \%classlist: A pointer to a hash containing the following data: -A list of student name:domain (as keys) (known below as $name) -A hash pointer for each student containing lastname, generation, firstname, middlename, and PID : Key is $name.studentInformation -A hash pointer to each students section data : Key is $name.section -If there was an error in dump, it will be returned in the hash. See the error codes for dump in lonnet. Also, an error key will be generated if an abort occurs. =back =cut sub DownloadClasslist { my ($courseID, $lastDownloadTime, $c)=@_; my ($courseDomain,$courseNumber)=split(/\_/,$courseID); my %classlist; my $modifiedTime = &Apache::lonnet::GetFileTimestamp($courseDomain, $courseNumber, 'classlist.db', $Apache::lonnet::perlvar{'lonUsersDir'}); # Always download the information if lastDownloadTime is set to # Not downloaded, otherwise it is only downloaded if the file # has been updated and has a more recent date stamp if($lastDownloadTime ne 'Not downloaded' && $lastDownloadTime >= $modifiedTime && $modifiedTime >= 0) { # Data is not gathered so return UpToDate as true. This # will be interpreted in ProcessClasslist $classlist{'lastDownloadTime'}=time; $classlist{'UpToDate'} = 'true'; return \%classlist; } %classlist=&Apache::lonnet::dump('classlist',$courseDomain, $courseNumber); foreach(keys (%classlist)) { if(/^(con_lost|error|no_such_host)/i) { return \%classlist; } } foreach my $name (keys(%classlist)) { if(defined($c) && ($c->aborted())) { $classlist{'error'}='aborted'; return \%classlist; } my ($studentName,$studentDomain) = split(/\:/,$name); # Download student environment data, specifically the full name and id. my %studentInformation=&Apache::lonnet::get('environment', ['lastname','generation', 'firstname','middlename', 'id'], $studentDomain, $studentName); $classlist{$name.':studentInformation'}=\%studentInformation; if($c->aborted()) { $classlist{'error'}='aborted'; return \%classlist; } #Section my %section=&Apache::lonnet::dump('roles',$studentDomain,$studentName); $classlist{$name.':sections'}=\%section; } $classlist{'UpToDate'} = 'false'; $classlist{'lastDownloadTime'}=time; return \%classlist; } =pod =item &DownloadCourseInformation() Dump of all the course information for a single student. The data can be pruned by making use of dumps regular expression arguement. This function also takes a regular expression which it passes straight through to dump. The data is no escaped, because it is done elsewhere. It also checks the timestamp of the students course database file and only downloads if it has been modified since the last download. =over 4 Input: $namedata, $courseID, $lastDownloadTime, $WhatIWant $namedata: student name:domain $courseID: The id of the course $lastDownloadTime: This is the date stamp for when this information was last gathered. If it is set to Not downloaded, it will gather the data again, though it currently does not remove the old data. $WhatIWant: Regular expression used to get selected data with dump Output: \%courseData \%courseData: A hash pointer to the raw data from the students course database. =back =cut sub DownloadCourseInformation { my ($namedata,$courseID,$lastDownloadTime,$WhatIWant)=@_; my %courseData; my ($name,$domain) = split(/\:/,$namedata); my $modifiedTime = &Apache::lonnet::GetFileTimestamp($domain, $name, $courseID.'.db', $Apache::lonnet::perlvar{'lonUsersDir'}); if($lastDownloadTime >= $modifiedTime && $modifiedTime >= 0) { # Data is not gathered so return UpToDate as true. This # will be interpreted in ProcessClasslist $courseData{$namedata.':lastDownloadTime'}=time; $courseData{$namedata.':UpToDate'} = 'true'; return \%courseData; } # Download course data if(!defined($WhatIWant)) { # set the regular expression to everything by setting it to period $WhatIWant = '.'; } %courseData=&Apache::lonnet::dump($courseID, $domain, $name, $WhatIWant); $courseData{'UpToDate'} = 'false'; $courseData{'lastDownloadTime'}=time; my %newData; foreach (keys(%courseData)) { # need to have the keys to be prepended with the name:domain of the # student to reduce data collision later. $newData{$namedata.':'.$_} = $courseData{$_}; } return \%newData; } # ----- END DOWNLOAD INFORMATION --------------------------------------- =pod =head1 PROCESSING FUNCTIONS These functions process all the data for all the students. Also, they are the functions that access the cache database for writing the majority of the time. The downloading and caching were separated to reduce problems with stopping downloading then can not tie hash to database later. =cut # ----- PROCESSING FUNCTIONS --------------------------------------- =pod =item &ProcessTopResourceMap() Trace through the "big hash" created in rat/lonuserstate.pm::loadmap. Basically, this function organizes a subset of the data and stores it in cached data. The data stored is the problems, sequences, sequence titles, parts of problems, and their ordering. Column width information is also partially handled here on a per sequence basis. =over 4 Input: $cache, $c $cache: A pointer to a hash to store the information $c: The connection class used to determine if an abort has been sent to the browser Output: A string that contains an error message or "OK" if everything went smoothly. =back =cut sub ProcessTopResourceMap { my ($cache,$c)=@_; my %hash; my $fn=$ENV{'request.course.fn'}; if(-e "$fn.db") { my $tieTries=0; while($tieTries < 3) { if($c->aborted()) { return; } if(tie(%hash,'GDBM_File',"$fn.db",&GDBM_READER(),0640)) { last; } $tieTries++; sleep 1; } if($tieTries >= 3) { return 'Coursemap undefined.'; } } else { return 'Can not open Coursemap.'; } my $oldkeys; if(defined($cache->{'ResourceKeys'})) { $oldkeys = $cache->{'ResourceKeys'}; foreach (split(':::', $cache->{'ResourceKeys'})) { delete $cache->{$_}; } delete $cache->{'ResourceKeys'}; } # Initialize state machine. Set information pointing to top level map. my (@sequences, @currentResource, @finishResource); my ($currentSequence, $currentResourceID, $lastResourceID); $currentResourceID=$hash{'ids_/res/'.$ENV{'request.course.uri'}}; push(@currentResource, $currentResourceID); $lastResourceID=-1; $currentSequence=-1; my $topLevelSequenceNumber = $currentSequence; my %sequenceRecord; my %allkeys; while(1) { if($c->aborted()) { last; } # HANDLE NEW SEQUENCE! #if page || sequence if(defined($hash{'map_pc_'.$hash{'src_'.$currentResourceID}}) && !defined($sequenceRecord{$currentResourceID})) { $sequenceRecord{$currentResourceID}++; push(@sequences, $currentSequence); push(@currentResource, $currentResourceID); push(@finishResource, $lastResourceID); $currentSequence=$hash{'map_pc_'.$hash{'src_'.$currentResourceID}}; # Mark sequence as containing problems. If it doesn't, then # it will be removed when processing for this sequence is # complete. This allows the problems in a sequence # to be outputed before problems in the subsequences if(!defined($cache->{'orderedSequences'})) { $cache->{'orderedSequences'}=$currentSequence; } else { $cache->{'orderedSequences'}.=':'.$currentSequence; } $allkeys{'orderedSequences'}++; $lastResourceID=$hash{'map_finish_'. $hash{'src_'.$currentResourceID}}; $currentResourceID=$hash{'map_start_'. $hash{'src_'.$currentResourceID}}; if(!($currentResourceID) || !($lastResourceID)) { $currentSequence=pop(@sequences); $currentResourceID=pop(@currentResource); $lastResourceID=pop(@finishResource); if($currentSequence eq $topLevelSequenceNumber) { last; } } next; } # Handle gradable resources: exams, problems, etc $currentResourceID=~/(\d+)\.(\d+)/; my $partA=$1; my $partB=$2; if($hash{'src_'.$currentResourceID}=~ /\.(problem|exam|quiz|assess|survey|form)$/ && $partA eq $currentSequence && !defined($sequenceRecord{$currentSequence.':'. $currentResourceID})) { $sequenceRecord{$currentSequence.':'.$currentResourceID}++; my $Problem = &Apache::lonnet::symbclean( &Apache::lonnet::declutter($hash{'map_id_'.$partA}). '___'.$partB.'___'. &Apache::lonnet::declutter($hash{'src_'. $currentResourceID})); $cache->{$currentResourceID.':problem'}=$Problem; $allkeys{$currentResourceID.':problem'}++; if(!defined($cache->{$currentSequence.':problems'})) { $cache->{$currentSequence.':problems'}=$currentResourceID; } else { $cache->{$currentSequence.':problems'}.= ':'.$currentResourceID; } $allkeys{$currentSequence.':problems'}++; my $meta=$hash{'src_'.$currentResourceID}; # $cache->{$currentResourceID.':title'}= # &Apache::lonnet::metdata($meta,'title'); $cache->{$currentResourceID.':title'}= $hash{'title_'.$currentResourceID}; $allkeys{$currentResourceID.':title'}++; $cache->{$currentResourceID.':source'}= $hash{'src_'.$currentResourceID}; $allkeys{$currentResourceID.':source'}++; # Get Parts for problem my %beenHere; foreach (split(/\,/,&Apache::lonnet::metadata($meta,'packages'))) { if(/^\w+response_\d+.*/) { my (undef, $partId, $responseId) = split(/_/,$_); if($beenHere{'p:'.$partId} == 0) { $beenHere{'p:'.$partId}++; if(!defined($cache->{$currentSequence.':'. $currentResourceID.':parts'})) { $cache->{$currentSequence.':'.$currentResourceID. ':parts'}=$partId; } else { $cache->{$currentSequence.':'.$currentResourceID. ':parts'}.=':'.$partId; } $allkeys{$currentSequence.':'.$currentResourceID. ':parts'}++; } if($beenHere{'r:'.$partId.':'.$responseId} == 0) { $beenHere{'r:'.$partId.':'.$responseId}++; if(!defined($cache->{$currentSequence.':'. $currentResourceID.':'.$partId. ':responseIDs'})) { $cache->{$currentSequence.':'.$currentResourceID. ':'.$partId.':responseIDs'}=$responseId; } else { $cache->{$currentSequence.':'.$currentResourceID. ':'.$partId.':responseIDs'}.=':'. $responseId; } $allkeys{$currentSequence.':'.$currentResourceID.':'. $partId.':responseIDs'}++; } if(/^optionresponse/ && $beenHere{'o:'.$partId.':'.$currentResourceID} == 0) { $beenHere{'o:'.$partId.$currentResourceID}++; if(defined($cache->{'OptionResponses'})) { $cache->{'OptionResponses'}.= ':::'. $currentSequence.':'.$currentResourceID.':'. $partId.':'.$responseId; } else { $cache->{'OptionResponses'}= $currentSequence.':'. $currentResourceID.':'. $partId.':'.$responseId; } $allkeys{'OptionResponses'}++; } } } } # if resource == finish resource, then it is the end of a sequence/page if($currentResourceID eq $lastResourceID) { # pop off last resource of sequence $currentResourceID=pop(@currentResource); $lastResourceID=pop(@finishResource); if(defined($cache->{$currentSequence.':problems'})) { # Capture sequence information here $cache->{$currentSequence.':title'}= $hash{'title_'.$currentResourceID}; $allkeys{$currentSequence.':title'}++; $cache->{$currentSequence.':source'}= $hash{'src_'.$currentResourceID}; $allkeys{$currentSequence.':source'}++; my $totalProblems=0; foreach my $currentProblem (split(/\:/, $cache->{$currentSequence. ':problems'})) { foreach (split(/\:/,$cache->{$currentSequence.':'. $currentProblem. ':parts'})) { $totalProblems++; } } my @titleLength=split(//,$cache->{$currentSequence. ':title'}); # $extra is 3 for problems correct and 3 for space # between problems correct and problem output my $extra = 6; if(($totalProblems + $extra) > (scalar @titleLength)) { $cache->{$currentSequence.':columnWidth'}= $totalProblems + $extra; } else { $cache->{$currentSequence.':columnWidth'}= (scalar @titleLength); } $allkeys{$currentSequence.':columnWidth'}++; } else { # Remove sequence from list, if it contains no problems to # display. $cache->{'orderedSequences'}=~s/$currentSequence//; $cache->{'orderedSequences'}=~s/::/:/g; $cache->{'orderedSequences'}=~s/^:|:$//g; } $currentSequence=pop(@sequences); if($currentSequence eq $topLevelSequenceNumber) { last; } } # MOVE!!! # move to next resource unless(defined($hash{'to_'.$currentResourceID})) { # big problem, need to handle. Next is probably wrong my $errorMessage = 'Big problem in '; $errorMessage .= 'loncoursedata::ProcessTopLevelMap.'; $errorMessage .= ' bighash to_$currentResourceID not defined!'; &Apache::lonnet::logthis($errorMessage); last; } my @nextResources=(); foreach (split(/\,/,$hash{'to_'.$currentResourceID})) { if(!defined($sequenceRecord{$currentSequence.':'. $hash{'goesto_'.$_}})) { push(@nextResources, $hash{'goesto_'.$_}); } } push(@currentResource, @nextResources); # Set the next resource to be processed $currentResourceID=pop(@currentResource); } my @theKeys = keys(%allkeys); my $newkeys = join(':::', @theKeys); $cache->{'ResourceKeys'} = join(':::', $newkeys); if($newkeys ne $oldkeys) { $cache->{'ResourceUpdated'} = 'true'; } else { $cache->{'ResourceUpdated'} = 'false'; } unless (untie(%hash)) { &Apache::lonnet::logthis("WARNING: ". "Could not untie coursemap $fn (browse)". "."); } return 'OK'; } =pod =item &ProcessClasslist() Taking the class list dumped from &DownloadClasslist(), all the students and their non-class information is processed using the &ProcessStudentInformation() function. A date stamp is also recorded for when the data was processed. Takes data downloaded for a student and breaks it up into managable pieces and stored in cache data. The username, domain, class related date, PID, full name, and section are all processed here. =over 4 Input: $cache, $classlist, $courseID, $ChartDB, $c $cache: A hash pointer to store the data $classlist: The hash of data collected about a student from &DownloadClasslist(). The hash contains a list of students, a pointer to a hash of student information for each student, and each students section number. $courseID: The course ID $ChartDB: The name of the cache database file. $c: The connection class used to determine if an abort has been sent to the browser Output: @names @names: An array of students whose information has been processed, and are to be considered in an arbitrary order. =back =cut sub ProcessClasslist { my ($cache,$classlist,$courseID,$c)=@_; my @names=(); $cache->{'ClasslistTimeStamp'}=$classlist->{'lastDownloadTime'}; if($classlist->{'UpToDate'} eq 'true') { return split(/:::/,$cache->{'NamesOfStudents'});; } foreach my $name (keys(%$classlist)) { if($name =~ /\:section/ || $name =~ /\:studentInformation/ || $name eq '' || $name eq 'UpToDate' || $name eq 'lastDownloadTime') { next; } if($c->aborted()) { return (); } my $studentInformation = $classlist->{$name.':studentInformation'}, my $sectionData = $classlist->{$name.':sections'}, my $date = $classlist->{$name}, my ($studentName,$studentDomain) = split(/\:/,$name); $cache->{$name.':username'}=$studentName; $cache->{$name.':domain'}=$studentDomain; # Initialize timestamp for student if(!defined($cache->{$name.':lastDownloadTime'})) { $cache->{$name.':lastDownloadTime'}='Not downloaded'; $cache->{$name.':updateTime'}=' Not updated'; } my $error = 0; foreach(keys(%$studentInformation)) { if(/^(con_lost|error|no_such_host)/i) { $cache->{$name.':error'}= 'Could not download student environment data.'; $cache->{$name.':fullname'}=''; $cache->{$name.':id'}=''; $error = 1; } } next if($error); push(@names,$name); $cache->{$name.':fullname'}=&ProcessFullName( $studentInformation->{'lastname'}, $studentInformation->{'generation'}, $studentInformation->{'firstname'}, $studentInformation->{'middlename'}); $cache->{$name.':id'}=$studentInformation->{'id'}; my ($end, $start)=split(':',$date); $courseID=~s/\_/\//g; $courseID=~s/^(\w)/\/$1/; my $sec=''; foreach my $key (keys (%$sectionData)) { my $value = $sectionData->{$key}; if ($key=~/^$courseID(?:\/)*(\w+)*\_st$/) { my $tempsection=$1; if($key eq $courseID.'_st') { $tempsection=''; } my ($dummy,$roleend,$rolestart)=split(/\_/,$value); if($roleend eq $end && $rolestart eq $start) { $sec = $tempsection; last; } } } my $status='Expired'; if(((!$end) || time < $end) && ((!$start) || (time > $start))) { $status='Active'; } $cache->{$name.':Status'}=$status; $cache->{$name.':section'}=$sec; if($sec eq '' || !defined($sec) || $sec eq ' ') { $sec = 'none'; } if(defined($cache->{'sectionList'})) { if($cache->{'sectionList'} !~ /(^$sec:|^$sec$|:$sec$|:$sec:)/) { $cache->{'sectionList'} .= ':'.$sec; } } else { $cache->{'sectionList'} = $sec; } } $cache->{'ClasslistTimestamp'}=time; $cache->{'NamesOfStudents'}=join(':::',@names); return @names; } =pod =item &ProcessStudentData() Takes the course data downloaded for a student in &DownloadCourseInformation() and breaks it up into key value pairs to be stored in the cached data. The keys are comprised of the $username:$domain:$keyFromCourseDatabase. The student username:domain is stored away signifying that the students information has been downloaded and can be reused from cached data. =over 4 Input: $cache, $courseData, $name $cache: A hash pointer to store data $courseData: A hash pointer that points to the course data downloaded for a student. $name: username:domain Output: None *NOTE: There is no output, but an error message is stored away in the cache data. This is checked in &FormatStudentData(). The key username:domain:error will only exist if an error occured. The error is an error from &DownloadCourseInformation(). =back =cut sub ProcessStudentData { my ($cache,$courseData,$name)=@_; if(!&CheckDateStampError($courseData, $cache, $name)) { return; } # This little delete thing, should not be here. Move some other # time though. if(defined($cache->{$name.':keys'})) { foreach (split(':::', $cache->{$name.':keys'})) { delete $cache->{$name.':'.$_}; } delete $cache->{$name.':keys'}; } my %courseKeys; # user name:domain was prepended earlier in DownloadCourseInformation foreach (keys %$courseData) { my $currentKey = $_; $currentKey =~ s/^$name//; $courseKeys{$currentKey}++; $cache->{$_}=$courseData->{$_}; } $cache->{$name.':keys'} = join(':::', keys(%courseKeys)); return; } =pod =item &ExtractStudentData() HISTORY: This function originally existed in every statistics module, and performed different tasks, the had some overlap. Due to the need for the data from the different modules, they were combined into a single function. This function now extracts all the necessary course data for a student from what was downloaded from their homeserver. There is some extra time overhead compared to the ProcessStudentInformation function, but it would have had to occurred at some point anyways. This is now typically called while downloading the data it will process. It is the brother function to ProcessStudentInformation. =over 4 Input: $input, $output, $data, $name $input: A hash that contains the input data to be processed $output: A hash to contain the processed data $data: A hash containing the information on what is to be processed and how (basically). $name: username:domain The input is slightly different here, but is quite simple. It is currently used where the $input, $output, and $data can and are often the same hashes, but they do not need to be. Output: None *NOTE: There is no output, but an error message is stored away in the cache data. This is checked in &FormatStudentData(). The key username:domain:error will only exist if an error occured. The error is an error from &DownloadCourseInformation(). =back =cut sub ExtractStudentData { my ($input, $output, $data, $name)=@_; if(!&CheckDateStampError($input, $data, $name)) { return; } # This little delete thing, should not be here. Move some other # time though. my %allkeys; if(defined($output->{$name.':keys'})) { foreach (split(':::', $output->{$name.':keys'})) { delete $output->{$name.':'.$_}; } delete $output->{$name.':keys'}; } my ($username,$domain)=split(':',$name); my $Version; my $problemsCorrect = 0; my $totalProblems = 0; my $problemsSolved = 0; my $numberOfParts = 0; my $totalAwarded = 0; foreach my $sequence (split(':', $data->{'orderedSequences'})) { foreach my $problemID (split(':', $data->{$sequence.':problems'})) { my $problem = $data->{$problemID.':problem'}; my $LatestVersion = $input->{$name.':version:'.$problem}; # Output dashes for all the parts of this problem if there # is no version information about the current problem. $output->{$name.':'.$problemID.':NoVersion'} = 'false'; $allkeys{$name.':'.$problemID.':NoVersion'}++; if(!$LatestVersion) { foreach my $part (split(/\:/,$data->{$sequence.':'. $problemID. ':parts'})) { $output->{$name.':'.$problemID.':'.$part.':tries'} = 0; $output->{$name.':'.$problemID.':'.$part.':awarded'} = 0; $output->{$name.':'.$problemID.':'.$part.':code'} = ' '; $allkeys{$name.':'.$problemID.':'.$part.':tries'}++; $allkeys{$name.':'.$problemID.':'.$part.':awarded'}++; $allkeys{$name.':'.$problemID.':'.$part.':code'}++; $totalProblems++; } $output->{$name.':'.$problemID.':NoVersion'} = 'true'; next; } my %partData=undef; # Initialize part data, display skips correctly # Skip refers to when a student made no submissions on that # part/problem. foreach my $part (split(/\:/,$data->{$sequence.':'. $problemID. ':parts'})) { $partData{$part.':tries'}=0; $partData{$part.':code'}=' '; $partData{$part.':awarded'}=0; $partData{$part.':timestamp'}=0; foreach my $response (split(':', $data->{$sequence.':'. $problemID.':'. $part.':responseIDs'})) { $partData{$part.':'.$response.':submission'}=''; } } # Looping through all the versions of each part, starting with the # oldest version. Basically, it gets the most recent # set of grade data for each part. my @submissions = (); for(my $Version=1; $Version<=$LatestVersion; $Version++) { foreach my $part (split(/\:/,$data->{$sequence.':'. $problemID. ':parts'})) { if(!defined($input->{"$name:$Version:$problem". ":resource.$part.solved"})) { # No grade for this submission, so skip next; } my $tries=0; my $code=' '; my $awarded=0; $tries = $input->{$name.':'.$Version.':'.$problem. ':resource.'.$part.'.tries'}; $awarded = $input->{$name.':'.$Version.':'.$problem. ':resource.'.$part.'.awarded'}; $partData{$part.':awarded'}=($awarded) ? $awarded : 0; $partData{$part.':tries'}=($tries) ? $tries : 0; $partData{$part.':timestamp'}=$input->{$name.':'.$Version.':'. $problem. ':timestamp'}; if(!$input->{$name.':'.$Version.':'.$problem.':resource.'.$part. '.previous'}) { foreach my $response (split(':', $data->{$sequence.':'. $problemID.':'. $part.':responseIDs'})) { @submissions=($input->{$name.':'.$Version.':'. $problem. ':resource.'.$part.'.'. $response.'.submission'}, @submissions); } } my $val = $input->{$name.':'.$Version.':'.$problem. ':resource.'.$part.'.solved'}; if ($val eq 'correct_by_student') {$code = '*';} elsif ($val eq 'correct_by_override') {$code = '+';} elsif ($val eq 'incorrect_attempted') {$code = '.';} elsif ($val eq 'incorrect_by_override'){$code = '-';} elsif ($val eq 'excused') {$code = 'x';} elsif ($val eq 'ungraded_attempted') {$code = '#';} else {$code = ' ';} $partData{$part.':code'}=$code; } } foreach my $part (split(/\:/,$data->{$sequence.':'.$problemID. ':parts'})) { $output->{$name.':'.$problemID.':'.$part.':wrong'} = $partData{$part.':tries'}; $allkeys{$name.':'.$problemID.':'.$part.':wrong'}++; if($partData{$part.':code'} eq '*') { $output->{$name.':'.$problemID.':'.$part.':wrong'}--; $problemsCorrect++; } elsif($partData{$part.':code'} eq '+') { $output->{$name.':'.$problemID.':'.$part.':wrong'}--; $problemsCorrect++; } $output->{$name.':'.$problemID.':'.$part.':tries'} = $partData{$part.':tries'}; $output->{$name.':'.$problemID.':'.$part.':code'} = $partData{$part.':code'}; $output->{$name.':'.$problemID.':'.$part.':awarded'} = $partData{$part.':awarded'}; $allkeys{$name.':'.$problemID.':'.$part.':tries'}++; $allkeys{$name.':'.$problemID.':'.$part.':code'}++; $allkeys{$name.':'.$problemID.':'.$part.':awarded'}++; $totalAwarded += $partData{$part.':awarded'}; $output->{$name.':'.$problemID.':'.$part.':timestamp'} = $partData{$part.':timestamp'}; $allkeys{$name.':'.$problemID.':'.$part.':timestamp'}++; foreach my $response (split(':', $data->{$sequence.':'. $problemID.':'. $part.':responseIDs'})) { $output->{$name.':'.$problemID.':'.$part.':'.$response. ':submission'}=join(':::',@submissions); $allkeys{$name.':'.$problemID.':'.$part.':'.$response. ':submission'}++; } if($partData{$part.':code'} ne 'x') { $totalProblems++; } } } $output->{$name.':'.$sequence.':problemsCorrect'} = $problemsCorrect; $allkeys{$name.':'.$sequence.':problemsCorrect'}++; $problemsSolved += $problemsCorrect; $problemsCorrect=0; } $output->{$name.':problemsSolved'} = $problemsSolved; $output->{$name.':totalProblems'} = $totalProblems; $output->{$name.':totalAwarded'} = $totalAwarded; $allkeys{$name.':problemsSolved'}++; $allkeys{$name.':totalProblems'}++; $allkeys{$name.':totalAwarded'}++; $output->{$name.':keys'} = join(':::', keys(%allkeys)); return; } sub LoadDiscussion { my ($courseID)=@_; my %Discuss=(); my %contrib=&Apache::lonnet::dump( $courseID, $ENV{'course.'.$courseID.'.domain'}, $ENV{'course.'.$courseID.'.num'}); #my %contrib=&DownloadCourseInformation($name, $courseID, 0); foreach my $temp(keys %contrib) { if ($temp=~/^version/) { my $ver=$contrib{$temp}; my ($dummy,$prb)=split(':',$temp); for (my $idx=1; $idx<=$ver; $idx++ ) { my $name=$contrib{"$idx:$prb:sendername"}; $Discuss{"$name:$prb"}=$idx; } } } return \%Discuss; } # ----- END PROCESSING FUNCTIONS --------------------------------------- =pod =head1 HELPER FUNCTIONS These are just a couple of functions do various odd and end jobs. There was also a couple of bulk functions added. These are &DownloadStudentCourseData(), &DownloadStudentCourseDataSeparate(), and &CheckForResidualDownload(). These functions now act as the interface for downloading student course data. The statistical modules should no longer make the calls to dump and download and process etc. They make calls to these bulk functions to get their data. =cut # ----- HELPER FUNCTIONS ----------------------------------------------- sub CheckDateStampError { my ($courseData, $cache, $name)=@_; if($courseData->{$name.':UpToDate'} eq 'true') { $cache->{$name.':lastDownloadTime'} = $courseData->{$name.':lastDownloadTime'}; if($courseData->{$name.':lastDownloadTime'} eq 'Not downloaded') { $cache->{$name.':updateTime'} = ' Not updated'; } else { $cache->{$name.':updateTime'}= localtime($courseData->{$name.':lastDownloadTime'}); } return 0; } $cache->{$name.':lastDownloadTime'}=$courseData->{$name.':lastDownloadTime'}; if($courseData->{$name.':lastDownloadTime'} eq 'Not downloaded') { $cache->{$name.':updateTime'} = ' Not updated'; } else { $cache->{$name.':updateTime'}= localtime($courseData->{$name.':lastDownloadTime'}); } if(defined($courseData->{$name.':error'})) { $cache->{$name.':error'}=$courseData->{$name.':error'}; return 0; } return 1; } =pod =item &ProcessFullName() Takes lastname, generation, firstname, and middlename (or some partial set of this data) and returns the full name version as a string. Format is Lastname generation, firstname middlename or a subset of this. =cut sub ProcessFullName { my ($lastname, $generation, $firstname, $middlename)=@_; my $Str = ''; if($lastname ne '') { $Str .= $lastname.' '; if($generation ne '') { $Str .= $generation; } else { chop($Str); } $Str .= ', '; if($firstname ne '') { $Str .= $firstname.' '; } if($middlename ne '') { $Str .= $middlename; } else { chop($Str); if($firstname eq '') { chop($Str); } } } else { if($firstname ne '') { $Str .= $firstname.' '; } if($middlename ne '') { $Str .= $middlename.' '; } if($generation ne '') { $Str .= $generation; } else { chop($Str); } } return $Str; } =pod =item &TestCacheData() Determine if the cache database can be accessed with a tie. It waits up to ten seconds before returning failure. This function exists to help with the problems with stopping the data download. When an abort occurs and the user quickly presses a form button and httpd child is created. This child needs to wait for the other to finish (hopefully within ten seconds). =over 4 Input: $ChartDB $ChartDB: The name of the cache database to be opened Output: -1, 0, 1 -1: Could not tie database 0: Use cached data 1: New cache database created, use that. =back =cut sub TestCacheData { my ($ChartDB,$isRecalculate,$totalDelay)=@_; my $isCached=-1; my %testData; my $tieTries=0; if(!defined($totalDelay)) { $totalDelay = 10; } if ((-e "$ChartDB") && (!$isRecalculate)) { $isCached = 1; } else { $isCached = 0; } while($tieTries < $totalDelay) { my $result=0; if($isCached) { $result=tie(%testData,'GDBM_File',$ChartDB,&GDBM_READER(),0640); } else { $result=tie(%testData,'GDBM_File',$ChartDB,&GDBM_NEWDB(),0640); } if($result) { last; } $tieTries++; sleep 1; } if($tieTries >= $totalDelay) { return -1; } untie(%testData); return $isCached; } sub DownloadStudentCourseData { my ($students,$checkDate,$cacheDB,$extract,$status,$courseID,$r,$c)=@_; my $title = 'LON-CAPA Statistics'; my $heading = 'Download and Process Course Data'; my $studentCount = scalar(@$students); my $WhatIWant; $WhatIWant = '(^version:|'; $WhatIWant .= '^\d+:.+?:(resource\.\d+\.'; $WhatIWant .= '(solved|tries|previous|awarded|(\d+\.submission))\s*$'; $WhatIWant .= '|timestamp)'; $WhatIWant .= ')'; # $WhatIWant = '.'; if($status eq 'true') { &Apache::lonhtmlcommon::Create_PrgWin($r, $title, $heading); } my $displayString; my $count=0; foreach (@$students) { my %cache; if($c->aborted()) { return 'Aborted'; } if($status eq 'true') { $count++; my $displayString = $count.'/'.$studentCount.': '.$_; &Apache::lonhtmlcommon::Update_PrgWin($displayString, $r); } my $downloadTime='Not downloaded'; my $needUpdate = 'false'; if($checkDate eq 'true' && tie(%cache,'GDBM_File',$cacheDB,&GDBM_READER(),0640)) { $downloadTime = $cache{$_.':lastDownloadTime'}; $needUpdate = $cache{'ResourceUpdated'}; untie(%cache); } if($c->aborted()) { return 'Aborted'; } if($needUpdate eq 'true') { $downloadTime = 'Not downloaded'; } my $courseData = &DownloadCourseInformation($_, $courseID, $downloadTime, $WhatIWant); if(tie(%cache,'GDBM_File',$cacheDB,&GDBM_WRCREAT(),0640)) { foreach my $key (keys(%$courseData)) { if($key =~ /^(con_lost|error|no_such_host)/i) { $courseData->{$_.':error'} = 'No course data for '.$_; last; } } if($extract eq 'true') { &ExtractStudentData($courseData, \%cache, \%cache, $_); } else { &ProcessStudentData(\%cache, $courseData, $_); } untie(%cache); } else { next; } } if($status eq 'true') { &Apache::lonhtmlcommon::Close_PrgWin($r); } return 'OK'; } sub DownloadStudentCourseDataSeparate { my ($students,$checkDate,$cacheDB,$extract,$status,$courseID,$r,$c)=@_; my $residualFile = '/home/httpd/perl/tmp/'.$courseID.'DownloadFile.db'; my $title = 'LON-CAPA Statistics'; my $heading = 'Download Course Data'; my $WhatIWant; $WhatIWant = '(^version:|'; $WhatIWant .= '^\d+:.+?:(resource\.\d+\.'; $WhatIWant .= '(solved|tries|previous|awarded|(\d+\.submission))\s*$'; $WhatIWant .= '|timestamp)'; $WhatIWant .= ')'; &CheckForResidualDownload($courseID, $cacheDB, $students, $c); my $studentCount = scalar(@$students); if($status eq 'true') { &Apache::lonhtmlcommon::Create_PrgWin($r, $title, $heading); } my $count=0; my $displayString=''; foreach (@$students) { if($c->aborted()) { return 'Aborted'; } if($status eq 'true') { $count++; $displayString = $count.'/'.$studentCount.': '.$_; &Apache::lonhtmlcommon::Update_PrgWin($displayString, $r); } my %cache; my $downloadTime='Not downloaded'; my $needUpdate = 'false'; if($checkDate eq 'true' && tie(%cache,'GDBM_File',$cacheDB,&GDBM_READER(),0640)) { $downloadTime = $cache{$_.':lastDownloadTime'}; $needUpdate = $cache{'ResourceUpdated'}; untie(%cache); } if($c->aborted()) { return 'Aborted'; } if($needUpdate eq 'true') { $downloadTime = 'Not downloaded'; } my $error = 0; my $courseData = &DownloadCourseInformation($_, $courseID, $downloadTime, $WhatIWant); my %downloadData; unless(tie(%downloadData,'GDBM_File',$residualFile, &GDBM_WRCREAT(),0640)) { return 'Failed to tie temporary download hash.'; } foreach my $key (keys(%$courseData)) { $downloadData{$key} = $courseData->{$key}; if($key =~ /^(con_lost|error|no_such_host)/i) { $error = 1; last; } } if($error) { foreach my $deleteKey (keys(%$courseData)) { delete $downloadData{$deleteKey}; } $downloadData{$_.':error'} = 'No course data for '.$_; } untie(%downloadData); } if($status eq 'true') { &Apache::lonhtmlcommon::Close_PrgWin($r); } return &CheckForResidualDownload($cacheDB, 'true', 'true', $courseID, $r, $c); } sub CheckForResidualDownload { my ($cacheDB,$extract,$status,$courseID,$r,$c)=@_; my $residualFile = '/home/httpd/perl/tmp/'.$courseID.'DownloadFile.db'; if(!-e $residualFile) { return 'OK'; } my %downloadData; my %cache; unless(tie(%downloadData,'GDBM_File',$residualFile,&GDBM_READER(),0640)) { return 'Can not tie database for check for residual download: tempDB'; } unless(tie(%cache,'GDBM_File',$cacheDB,&GDBM_WRCREAT(),0640)) { untie(%downloadData); return 'Can not tie database for check for residual download: cacheDB'; } my @students=(); my %checkStudent; my $key; while(($key, undef) = each %downloadData) { my @temp = split(':', $key); my $student = $temp[0].':'.$temp[1]; if(!defined($checkStudent{$student})) { $checkStudent{$student}++; push(@students, $student); } } my $heading = 'Process Course Data'; my $title = 'LON-CAPA Statistics'; my $studentCount = scalar(@students); if($status eq 'true') { &Apache::lonhtmlcommon::Create_PrgWin($r, $title, $heading); } my $count=1; foreach my $name (@students) { last if($c->aborted()); if($status eq 'true') { my $displayString = $count.'/'.$studentCount.': '.$name; &Apache::lonhtmlcommon::Update_PrgWin($displayString, $r); } if($extract eq 'true') { &ExtractStudentData(\%downloadData, \%cache, \%cache, $name); } else { &ProcessStudentData(\%cache, \%downloadData, $name); } $count++; } if($status eq 'true') { &Apache::lonhtmlcommon::Close_PrgWin($r); } untie(%cache); untie(%downloadData); if(!$c->aborted()) { my @files = ($residualFile); unlink(@files); } return 'OK'; } # ----- END HELPER FUNCTIONS -------------------------------------------- 1; __END__