--- loncom/interface/loncoursedata.pm 2002/07/17 12:38:25 1.2 +++ loncom/interface/loncoursedata.pm 2002/09/03 02:22:40 1.29 @@ -1,7 +1,7 @@ # The LearningOnline Network with CAPA # (Publication Handler # -# $Id: loncoursedata.pm,v 1.2 2002/07/17 12:38:25 stredwic Exp $ +# $Id: loncoursedata.pm,v 1.29 2002/09/03 02:22:40 albertel Exp $ # # Copyright Michigan State University Board of Trustees # @@ -35,12 +35,13 @@ loncoursedata =head1 SYNOPSIS -Set of functions that download and process student information. +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 @@ -51,6 +52,7 @@ package Apache::loncoursedata; use strict; use Apache::Constants qw(:common :http); use Apache::lonnet(); +use Apache::lonhtmlcommon; use HTML::TokeParser; use GDBM_File; @@ -58,11 +60,8 @@ use GDBM_File; =head1 DOWNLOAD INFORMATION -This section contains all the files that get data from other servers -and/or itself. There is one function that has a call to get remote -information but isn't included here which is ProcessTopLevelMap. The -usage was small enough to be ignored, but that portion may be moved -here in the future. +This section contains all the functions that get data from other servers +and/or itself. =cut @@ -70,20 +69,31 @@ here in the future. =pod -=item &DownloadNamePIDSection() +=item &DownloadClasslist() Collects lastname, generation, middlename, firstname, PID, and section for each -student from their environment database. The list of students is built from -collecting a classlist for the course that is to be displayed. +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, $c +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 doesn't continue to +is used to short circuit this function so that it does not continue to get information when there is no need. Output: \%classlist @@ -93,27 +103,48 @@ Output: \%classlist -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' +middlename, and PID : Key is $name.studentInformation -A hash pointer to each students section data : Key is $name.section +-If there was an error in dump, it will be returned in the hash. See +the error codes for dump in lonnet. Also, an error key will be +generated if an abort occurs. + =back =cut -sub DownloadStudentNamePIDSection { - my ($courseID, $c)=@_; +sub DownloadClasslist { + my ($courseID, $lastDownloadTime, $c)=@_; my ($courseDomain,$courseNumber)=split(/\_/,$courseID); + my %classlist; - my %classlist=&Apache::lonnet::dump('classlist',$courseDomain, - $courseNumber); - my ($checkForError)=keys (%classlist); - if($checkForError =~ /^(con_lost|error|no_such_host)/i) { + 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 $name (keys(%classlist)) { - if($c->aborted()) { + if(defined($c) && ($c->aborted())) { $classlist{'error'}='aborted'; return \%classlist; } @@ -135,44 +166,83 @@ sub DownloadStudentNamePIDSection { #Section my %section=&Apache::lonnet::dump('roles',$studentDomain,$studentName); - $classlist{$name.':section'}=\%section; + $classlist{$name.':sections'}=\%section; } + $classlist{'UpToDate'} = 'false'; + $classlist{'lastDownloadTime'}=time; + return \%classlist; } =pod -=item &DownloadStudentCourseInformation() +=item &DownloadCourseInformation() -Dump of all the course information for a single student. There is no -pruning of data, it is all stored in a hash and returned. +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: $name, $courseID +Input: $namedata, $courseID, $lastDownloadTime, $WhatIWant -$name: student name:domain +$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. + +$WhatIWant: Regular expression used to get selected data with dump + Output: \%courseData -\%courseData: A hash pointer to the raw data from the student's course +\%courseData: A hash pointer to the raw data from the students course database. =back =cut -sub DownloadStudentCourseInformation { - my ($name,$courseID)=@_; - my ($studentName,$studentDomain) = split(/\:/,$name); - - # Download student course data - my %courseData=&Apache::lonnet::dump($courseID,$studentDomain, - $studentName); - return \%courseData; +sub DownloadCourseInformation { + my ($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 --------------------------------------- @@ -182,10 +252,9 @@ sub DownloadStudentCourseInformation { =head1 PROCESSING FUNCTIONS These functions process all the data for all the students. Also, they -are the only functions that access the cache database for writing. Thus -they are the only functions that cache data. The downloading and caching -were separated to reduce problems with stopping downloading then can't -tie hash to database later. +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 @@ -227,7 +296,7 @@ sub ProcessTopResourceMap { if($c->aborted()) { return; } - if(tie(%hash,'GDBM_File',"$fn.db",&GDBM_READER,0640)) { + if(tie(%hash,'GDBM_File',"$fn.db",&GDBM_READER(),0640)) { last; } $tieTries++; @@ -240,6 +309,15 @@ sub ProcessTopResourceMap { 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); @@ -250,13 +328,17 @@ sub ProcessTopResourceMap { $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_'.$currentResourceID}})) { + if(defined($hash{'map_pc_'.$hash{'src_'.$currentResourceID}}) && + !defined($sequenceRecord{$currentResourceID})) { + $sequenceRecord{$currentResourceID}++; push(@sequences, $currentSequence); push(@currentResource, $currentResourceID); push(@finishResource, $lastResourceID); @@ -272,6 +354,7 @@ sub ProcessTopResourceMap { } else { $cache->{'orderedSequences'}.=':'.$currentSequence; } + $allkeys{'orderedSequences'}++; $lastResourceID=$hash{'map_finish_'. $hash{'src_'.$currentResourceID}}; @@ -286,6 +369,7 @@ sub ProcessTopResourceMap { last; } } + next; } # Handle gradable resources: exams, problems, etc @@ -294,7 +378,10 @@ sub ProcessTopResourceMap { my $partB=$2; if($hash{'src_'.$currentResourceID}=~ /\.(problem|exam|quiz|assess|survey|form)$/ && - $partA eq $currentSequence) { + $partA eq $currentSequence && + !defined($sequenceRecord{$currentSequence.':'. + $currentResourceID})) { + $sequenceRecord{$currentSequence.':'.$currentResourceID}++; my $Problem = &Apache::lonnet::symbclean( &Apache::lonnet::declutter($hash{'map_id_'.$partA}). '___'.$partB.'___'. @@ -302,50 +389,75 @@ sub ProcessTopResourceMap { $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 - foreach (split(/\,/,&Apache::lonnet::metadata($meta,'keys'))) { - if($_=~/^stores\_(\d+)\_tries$/) { - my $Part=&Apache::lonnet::metadata($meta,$_.'.part'); - if(!defined($cache->{$currentSequence.':'. - $currentResourceID.':parts'})) { - $cache->{$currentSequence.':'.$currentResourceID. - ':parts'}=$Part; - } else { - $cache->{$currentSequence.':'.$currentResourceID. - ':parts'}.=':'.$Part; + 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.':'.$currentResourceID. + ':parts'}++; } - foreach (split(/\,/, - &Apache::lonnet::metadata($meta,'packages'))) { - if($_=~/^optionresponse\_($Part)\_(\w+)$/) { - if(defined($cache->{'OptionResponses'})) { - $cache->{'OptionResponses'}.= ':::'. - $hash{'src_'.$currentResourceID}.'::'. - $hash{'title_'.$currentResourceID}.'::'. - $Part.'::'.$Problem; - } else { - $cache->{'OptionResponses'}= - $hash{'src_'.$currentResourceID}.'::'. - $hash{'title_'.$currentResourceID}.'::'. - $Part.'::'.$Problem; - } + 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) { @@ -357,8 +469,10 @@ sub ProcessTopResourceMap { # 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(/\:/, @@ -382,6 +496,7 @@ sub ProcessTopResourceMap { $cache->{$currentSequence.':columnWidth'}= (scalar @titleLength); } + $allkeys{$currentSequence.':columnWidth'}++; } else { # Remove sequence from list, if it contains no problems to # display. @@ -394,23 +509,39 @@ sub ProcessTopResourceMap { 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})) { - push(@nextResources, $hash{'goesto_'.$_}); + 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)". @@ -422,270 +553,455 @@ sub ProcessTopResourceMap { =pod -=item &ProcessSection() +=item &ProcessClasslist() -Determine the section number for a student for the class. A student can have -multiple sections for the same class. The correct one is chosen. +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: $sectionData, $courseid, $ActiveFlag +Input: $cache, $classlist, $courseID, $ChartDB, $c -$sectionData: A pointer to a hash containing all section data for this -student for the class +$cache: A hash pointer to store the data -$courseid: The course ID. +$classlist: The hash of data collected about a student from +&DownloadClasslist(). The hash contains a list of students, a pointer +to a hash of student information for each student, and each students section +number. -$ActiveFlag: The student's active status (Active/Expired) +$courseID: The course ID -Output: $oldsection, $cursection, or -1 +$ChartDB: The name of the cache database file. + +$c: The connection class used to determine if an abort has been sent to the +browser -$oldsection and $cursection and sections number that will be displayed in the -chart. +Output: @names --1 is returned if an error occurs. +@names: An array of students whose information has been processed, and are to +be considered in an arbitrary order. =back =cut -sub ProcessSection { - my ($sectionData,$courseid,$ActiveFlag)=@_; - $courseid=~s/\_/\//g; - $courseid=~s/^(\w)/\/$1/; - - my $cursection='-1'; - my $oldsection='-1'; - my $status='Expired'; - my $section=''; - foreach my $key (keys (%$sectionData)) { - my $value = $sectionData->{$key}; - if ($key=~/^$courseid(?:\/)*(\w+)*\_st$/) { - $section=$1; - if($key eq $courseid.'_st') { - $section=''; - } +sub ProcessClasslist { + my ($cache,$classlist,$courseID,$c)=@_; + my @names=(); - my ($dummy,$end,$start)=split(/\_/,$value); - my $now=time; - my $notactive=0; - if ($start) { - if($now<$start) { - $notactive=1; - } + $cache->{'ClasslistTimeStamp'}=$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'}=''; + $error = 1; } - if($end) { - if ($now>$end) { - $notactive=1; + } + next if($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 = $tempsection; + last; } } - if($notactive == 0) { - $status='Active'; - $cursection=$section; - last; - } - if($notactive == 1) { - $oldsection=$section; - } - } - } - if($status eq $ActiveFlag) { - if($cursection eq '-1') { - return $oldsection; } - return $cursection; - } - if($ActiveFlag eq 'Any') { - if($cursection eq '-1') { - return $oldsection; + + 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; } - return $cursection; } - return '-1'; + + $cache->{'ClasslistTimestamp'}=time; + $cache->{'NamesOfStudents'}=join(':::',@names); + + return @names; } =pod -=item &ProcessNamePIDSection() +=item &ProcessStudentData() -Takes data downloaded for a student and breaks it up into managable pieces and -stored in cache data. The username, domain, class related date, PID, -full name, and section are all processed here. +Takes the course data downloaded for a student in +&DownloadCourseInformation() and breaks it up into key value pairs +to be stored in the cached data. The keys are comprised of the +$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, $studentInformation, $section, $date, $name, $courseID - -$cache: A hash pointer to store the data - -$studentInformation: Student information is what was requested in -&DownloadPrerequistedData(). See that function for what data is requested. - -$section: A hash pointer to class section related information. +Input: $cache, $courseData, $name -$date: A composite of the start and end date for this class for this -student. Format: end:start +$cache: A hash pointer to store data -$name: the username:domain information +$courseData: A hash pointer that points to the course data downloaded for a +student. -$courseID: The course ID +$name: username:domain Output: None -*NOTE: There is no return value, but if an error occurs a key is added to -the cache data with the value being the error message. The key is -username:domain:error. It will only exist if an error occurs. +*NOTE: There is no output, but an error message is stored away in the cache +data. This is checked in &FormatStudentData(). The key username:domain:error +will only exist if an error occured. The error is an error from +&DownloadCourseInformation(). =back =cut -sub ProcessStudentNamePIDSection { - my ($cache,$studentInformation,$section,$date,$name,$courseID,$status)=@_; - my ($studentName,$studentDomain) = split(/\:/,$name); - - $cache->{$name.':username'}=$studentName; - $cache->{$name.':domain'}=$studentDomain; - $cache->{$name.':date'}=$date; - - my ($checkForError)=keys(%$studentInformation); - if($checkForError =~ /^(con_lost|error|no_such_host)/i) { - $cache->{$name.':error'}= - 'Could not download student environment data.'; - $cache->{$name.':fullname'}=''; - $cache->{$name.':id'}=''; - } else { - $cache->{$name.':fullname'}=&ProcessFullName( - $studentInformation->{'lastname'}, - $studentInformation->{'generation'}, - $studentInformation->{'firstname'}, - $studentInformation->{'middlename'}); - $cache->{$name.':id'}=$studentInformation->{'id'}; +sub ProcessStudentData { + my ($cache,$courseData,$name)=@_; + + if(!&CheckDateStampError($courseData, $cache, $name)) { + return; } - my $sec=&ProcessSection($section, $courseID, $status); - if($sec != -1) { - $cache->{$name.':section'}=$sec; - } else { - $cache->{$name.':section'}=''; + # 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 &ProcessClassList() +=item &ExtractStudentData() -Taking the class list dumped from &DownloadPrerequisiteData(), 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. +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: $cache, $classlist, $courseID, $ChartDB, $c +Input: $input, $output, $data, $name -$cache: A hash pointer to store the data +$input: A hash that contains the input data to be processed -$classlist: The hash of data collected about a student from -&DownloadPrerequisteData(). The hash contains a list of students, a pointer -to a hash of student information for each student, and each student's section -number. +$output: A hash to contain the processed data -$courseID: The course ID +$data: A hash containing the information on what is to be +processed and how (basically). -$ChartDB: The name of the cache database file. +$name: username:domain -$c: The connection class used to determine if an abort has been sent to the -browser +The input is slightly different here, but is quite simple. +It is currently used where the $input, $output, and $data +can and are often the same hashes, but they do not need +to be. -Output: @names +Output: None -@names: An array of students whose information has been processed, and are to -be considered in an arbitrary order. +*NOTE: There is no output, but an error message is stored away in the cache +data. This is checked in &FormatStudentData(). The key username:domain:error +will only exist if an error occured. The error is an error from +&DownloadCourseInformation(). =back =cut -sub ProcessClassList { - my ($cache,$classlist,$courseID,$status,$c)=@_; - my @names=(); +sub ExtractStudentData { + my ($input, $output, $data, $name)=@_; - foreach my $name (keys(%$classlist)) { - if($name =~ /\:section/ || $name =~ /\:studentInformation/ || - $name eq '') { - next; - } - if($c->aborted()) { - last; - } - push(@names,$name); - &ProcessStudentNamePIDSection($cache, - $classlist->{$name.':studentInformation'}, - $classlist->{$name.':section'}, - $classlist->{$name}, - $name,$courseID,$status); + if(!&CheckDateStampError($input, $data, $name)) { + return; } - # Time of download - $cache->{'time'}=localtime(); - - return @names; -} + # This little delete thing, should not be here. Move some other + # time though. + my %allkeys; + if(defined($output->{$name.':keys'})) { + foreach (split(':::', $output->{$name.':keys'})) { + delete $output->{$name.':'.$_}; + } + delete $output->{$name.':keys'}; + } -=pod + my ($username,$domain)=split(':',$name); -=item &ProcessStudentData() + 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.':'.$problemID.':NoVersion'} = 'true'; + next; + } -Takes the course data downloaded for a student in -&DownloadStudentCourseInformation() and breaks it up into key value pairs -to be stored in the cached data. The keys are comprised of the -$username:$domain:$keyFromCourseDatabase. The student username:domain is -stored away signifying that the student's information has been downloaded and -can be reused from cached data. + 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'}=''; + } + } -=over 4 + # 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; + } -Input: $cache, $courseData, $name + 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.'.'. + $response.'.submission'}, + @submissions); + } + } -$cache: A hash pointer to store data + 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'}=$code; + } + } -$courseData: A hash pointer that points to the course data downloaded for a -student. + 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'}--; + $problemsCorrect++; + } -$name: username:domain + $output->{$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.':'.$response. + ':submission'}++; + } -Output: None + if($partData{$part.':code'} ne 'x') { + $totalProblems++; + } + } + } -*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 -&DownloadStudentCourseInformation(). + $output->{$name.':'.$sequence.':problemsCorrect'} = $problemsCorrect; + $allkeys{$name.':'.$sequence.':problemsCorrect'}++; + $problemsSolved += $problemsCorrect; + $problemsCorrect=0; + } -=back + $output->{$name.':problemsSolved'} = $problemsSolved; + $output->{$name.':totalProblems'} = $totalProblems; + $output->{$name.':totalAwarded'} = $totalAwarded; + $allkeys{$name.':problemsSolved'}++; + $allkeys{$name.':totalProblems'}++; + $allkeys{$name.':totalAwarded'}++; -=cut + $output->{$name.':keys'} = join(':::', keys(%allkeys)); -sub ProcessStudentData { - my ($cache,$courseData,$name)=@_; + return; +} - my ($checkForError) = keys(%$courseData); - if($checkForError =~ /^(con_lost|error|no_such_host)/i) { - $cache->{$name.':error'}='Could not download course data.'; - } else { - foreach my $key (keys (%$courseData)) { - $cache->{$name.':'.$key}=$courseData->{$key}; - } - if(defined($cache->{'NamesOfStudents'})) { - $cache->{'NamesOfStudents'}.=':::'.$name; - } else { - $cache->{'NamesOfStudents'}=$name; - } - } +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; + return \%Discuss; } # ----- END PROCESSING FUNCTIONS --------------------------------------- @@ -695,12 +1011,47 @@ sub ProcessStudentData { =head1 HELPER FUNCTIONS These are just a couple of functions do various odd and end -jobs. +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() @@ -769,7 +1120,7 @@ $ChartDB: The name of the cache database Output: -1, 0, 1 --1: Couldn't tie database +-1: Could not tie database 0: Use cached data 1: New cache database created, use that. @@ -796,9 +1147,9 @@ sub TestCacheData { while($tieTries < $totalDelay) { my $result=0; if($isCached) { - $result=tie(%testData,'GDBM_File',$ChartDB,&GDBM_READER,0640); + $result=tie(%testData,'GDBM_File',$ChartDB,&GDBM_READER(),0640); } else { - $result=tie(%testData,'GDBM_File',$ChartDB,&GDBM_NEWDB,0640); + $result=tie(%testData,'GDBM_File',$ChartDB,&GDBM_NEWDB(),0640); } if($result) { last; @@ -815,34 +1166,223 @@ sub TestCacheData { return $isCached; } -#sub CheckStatus { -# my ($name, $data, $status)=@_; +sub DownloadStudentCourseData { + my ($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, $student); + } + } -# if($status eq 'Any') { -# my $section = ' '; -# foreach (split(':',$data->{$name.':Sections'})) { -# if($data->{$name.':'.$_.'Status'} eq 'Active') { -# return $_; -# } -# $section = $_; -# } -# return $_; -# } - -# foreach (split(':',$data->{$name.':Sections'})) { -# if($data->{$name.':'.$_.'Status'} eq $status) { -# return $_; -# } -# } - -# foreach (split(':',$data->{$name.':Sections'})) { -# if($data->{$name.':'.$_.'Status'} eq 'Any') { -# return $_; -# } -# } + 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); + } -# return 'not found'; -#} + 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 --------------------------------------------