[BACK]Return to loncoursedata.pm CVS log [TXT] [DIR] Up to [LON-CAPA] / loncom / interface

Diff for /loncom/interface/loncoursedata.pm between versions 1.183 and 1.195

version 1.183, 2007/10/03 18:22:50 version 1.195, 2013/12/09 21:33:32
Line 25 Line 25
 # http://www.lon-capa.org/  # http://www.lon-capa.org/
 #  #
 ###  ###
   
 =pod  =pod
   
 =head1 NAME  =head1 NAME
   
 loncoursedata  Apache::loncoursedata
   
 =head1 SYNOPSIS  =head1 SYNOPSIS
   
Line 44  Set of functions that download and proce Line 43  Set of functions that download and proce
   Apache::lonmysql    Apache::lonmysql
   LONCAPA    LONCAPA
   Digest::MD5    Digest::MD5
    
 =cut  =cut
   
 package Apache::loncoursedata;  package Apache::loncoursedata;
Line 57  use Apache::lonmysql(); Line 56  use Apache::lonmysql();
 use LONCAPA;  use LONCAPA;
 use Digest::MD5();  use Digest::MD5();
   
 =pod  =pod 
   
 =head1 DOWNLOAD INFORMATION  
   
 This section contains all the functions that get data from other servers   =head2 make_into_hash
 and/or itself.  
   
 =cut  
   
 ################################################  Turn a colon separated string into a hash and return a reference
 ################################################  to it.  Numbering from 0 even elements are keys and odd elements
   are values e.g. a:b:c:d creates a hash like
 =pod    a => b, c =>d
   
 =item &make_into_hash($values);  
   
 Returns a reference to a hash as described by $values.  $values is  
 assumed to be the result of   
     join(':',map {&escape($_)} %orighash);  
   
 This is a helper function for get_current_state.  
   
 =cut  =cut
   
 ################################################  
 ################################################  
 sub make_into_hash {  sub make_into_hash {
     my $values = shift;      my $values = shift;
     my %tmp = map { &unescape($_); } split(':',$values);      my %tmp = map { &unescape($_); } split(':',$values);
Line 90  sub make_into_hash { Line 74  sub make_into_hash {
 }  }
   
   
 ################################################  
 ################################################  
   
 =pod  
   
 =head1 LOCAL DATA CACHING SUBROUTINES  
   
 The local caching is done using MySQL.  There is no fall-back implementation  
 if MySQL is not running.  
   
 The programmers interface is to call &get_current_state() or some other  
 primary interface subroutine (described below).  The internals of this   
 storage system are documented here.  
   
 There are six tables used to store student performance data (the results of  
 a dumpcurrent).  Each of these tables is created in MySQL with a name of  
 $courseid_*****, where ***** is 'symb', 'part', or whatever is appropriate   
 for the table.  The tables and their purposes are described below.  
   
 Some notes before we get started.  
   
 Each table must have a PRIMARY KEY, which is a column or set of columns which  
 will serve to uniquely identify a row of data.  NULL is not allowed!  
   
 INDEXes work best on integer data.  
   
 JOIN is used to combine data from many tables into one output.  
   
 lonmysql.pm is used for some of the interface, specifically the table creation  
 calls.  The inserts are done in bulk by directly calling the database handler.  
 The SELECT ... JOIN statement used to retrieve the data does not have an  
 interface in lonmysql.pm and I shudder at the thought of writing one.  
   
 =head3 Table Descriptions  
   
 =over 4  
   
 =item Tables used to store meta information  
   
 The following tables hold data required to keep track of the current status  
 of a students data in the tables or to look up the students data in the tables.  
   
 =over 4  
   
 =item $symb_table  
   
 The symb_table has two columns.  The first is a 'symb_id' and the second  
 is the text name for the 'symb' (limited to 64k).  The 'symb_id' is generated  
 automatically by MySQL so inserts should be done on this table with an  
 empty first element.  This table has its PRIMARY KEY on the 'symb_id'.  
   
 =item $part_table  
   
 The part_table has two columns.  The first is a 'part_id' and the second  
 is the text name for the 'part' (limited to 100 characters).  The 'part_id' is  
 generated automatically by MySQL so inserts should be done on this table with  
 an empty first element.  This table has its PRIMARY KEY on the 'part' (100  
 characters) and a KEY on 'part_id'.  
   
 =item $student_table  
   
 The student_table has 7 columns.  The first is a 'student_id' assigned by   
 MySQL.  The second is 'student' which is username:domain.  The third through  
 fifth are 'section', 'status' (enrollment status), and 'classification'   
 (to be used in the future).  The sixth and seventh ('updatetime' and   
 'fullupdatetime') contain the time of last update and full update of student  
 data.  This table has its PRIMARY KEY on the 'student_id' column and is indexed  
 on 'student', 'section', and 'status'.  
   
 =item $groupnames_table  
   
 The groupnames_table has 2 columns.  The first is a 'group_id' assigned by   
 MySQL.  The second is 'groupname' which is the name of the group in the course.  
   
 =item $students_groups_table  
   
 The students_groups_table has 2 columns.  The first is the 'student_id', and the   
 second is the 'group_id'. These two columns comprise the PRIMARY KEY for this   
 table, as an individual student may be affiliated with more than one group at  
 any time. This table is indexed on both student_id and group_id.  
   
 =back   
   
 =item Tables used to store current status data  
   
 The following tables store data only about the students current status on   
 a problem, meaning only the data related to the last attempt on a problem.  
   
 =over 4  
   
 =item $performance_table  
   
 The performance_table has 9 columns.  The first three are 'symb_id',   
 'student_id', and 'part_id'.  These comprise the PRIMARY KEY for this table  
 and are directly related to the $symb_table, $student_table, and $part_table  
 described above.  MySQL does better indexing on numeric items than text,  
 so we use these three "index tables".  The remaining columns are  
 'solved', 'tries', 'awarded', 'award', 'awarddetail', and 'timestamp'.  
 These are either the MySQL type TINYTEXT or various integers ('tries' and   
 'timestamp').  This table has KEYs of 'student_id' and 'symb_id'.  
 For use of this table, see the functions described below.  
   
 =item $parameters_table  
   
 The parameters_table holds the data that does not fit neatly into the  
 performance_table.  The parameters table has four columns: 'symb_id',  
 'student_id', 'parameter', and 'value'.  'symb_id', 'student_id', and  
 'parameter' comprise the PRIMARY KEY for this table.  'parameter' is   
 limited to 255 characters.  'value' is limited to 64k characters.  
   
 =back  
   
 =item Tables used for storing historic data  
   
 The following tables are used to store almost all of the transactions a student  
 has made on a homework problem.  See loncapa/docs/homework/datastorage for   
 specific information about each of the parameters stored.    
   
 =over 4  
   
 =item $fulldump_response_table  
   
 The response table holds data (documented in loncapa/docs/homework/datastorage)  
 associated with a particular response id which is stored when a student   
 attempts a problem.  The following are the columns of the table, in order:  
 'symb_id','part_id','response_id','student_id','transaction','tries',  
 'awarddetail', 'response_specific', 'response_specific_value',  
 'response_specific_2', 'response_specific_value_2', and 'submission  
 (the text of the students submission).  The primary key is based on the  
 first five columns listed above.  
   
 =item $fulldump_part_table  
   
 The part table holds data (documented in loncapa/docs/homework/datastorage)  
 associated with a particular part id which is stored when a student attempts  
 a problem.  The following are the columns of the table, in order:  
 'symb_id','part_id','student_id','transaction','tries','award','awarded',  
 and 'previous'.  The primary key is based on the first five columns listed   
 above.  
   
 =item $fulldump_timestamp_table  
   
 The timestamp table holds the timestamps of the transactions which are  
 stored in $fulldump_response_table and $fulldump_part_table.  This data is  
 about both the response and part data.  Columns: 'symb_id','student_id',  
 'transaction', and 'timestamp'.    
 The primary key is based on the first 3 columns.  
   
 =item $weight_table  
   
 The weight table holds the weight for the problems used in the class.  
 Whereas the weight of a problem can vary by section and student the data  
 here is applied to the class as a whole.  
 Columns: 'symb_id','part_id','response_id','weight'.  
   
 =back  
   
 =back  
   
 =head3 Important Subroutines  
   
 Here is a brief overview of the subroutines which are likely to be of   
 interest:  
   
 =over 4  
   
 =item &get_current_state(): programmers interface.  
   
 =item &init_dbs(): table creation  
   
 =item &update_student_data(): data storage calls  
   
 =item &get_student_data_from_performance_cache(): data retrieval  
   
 =back  
   
 =head3 Main Documentation  
   
 =over 4  
   
 =cut  
   
 ################################################  
 ################################################  
   
 ################################################  
 ################################################  
 { # Begin scope of table identifiers  { # Begin scope of table identifiers
   
 my $current_course ='';  my $current_course ='';
Line 293  my $fulldump_timestamp_table; Line 90  my $fulldump_timestamp_table;
 my $weight_table;  my $weight_table;
   
 my @Tables;  my @Tables;
 ################################################  
 ################################################  
   
 =pod  
   
 =item &init_dbs()  
   
 Input: course id  
   
 Output: 0 on success, positive integer on error  
   
 This routine issues the calls to lonmysql to create the tables used to  
 store student data.  
   
 =cut  
   
 ################################################  
 ################################################  
 sub init_dbs {  sub init_dbs {
     my ($courseid,$drop) = @_;      my ($courseid,$drop) = @_;
     &setup_table_names($courseid);      &setup_table_names($courseid);
Line 665  sub init_dbs { Line 447  sub init_dbs {
     return 0;      return 0;
 }  }
   
 ################################################  
 ################################################  
   
 =pod  =pod
   
 =item &delete_caches()  =head2 delete_caches
   
 This routine drops all the tables associated with a course from the   Drops all of the tables in the local mysql cache associated with the
 MySQL database.  specified course id.
   
 Input: course id (optional, determined by environment if omitted)   
   
 Returns: nothing  TODO:  The drops should be pushed into lonmysql to further isolate 
   mysql code from other modules.
   
 =cut  =cut
   
 ################################################  
 ################################################  
 sub delete_caches {  sub delete_caches {
     my $courseid = shift;      my $courseid = shift;
     $courseid = $env{'request.course.id'} if (! defined($courseid));      $courseid = $env{'request.course.id'} if (! defined($courseid));
Line 700  sub delete_caches { Line 475  sub delete_caches {
     return;      return;
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &get_part_id()  
   
 Get the MySQL id of a problem part string.  
   
 Input: $part  
   
 Output: undef on error, integer $part_id on success.  
   
 =item &get_part()  
   
 Get the string describing a part from the MySQL id of the problem part.  
   
 Input: $part_id  
   
 Output: undef on error, $part string on success.  
   
 =cut  
   
 ################################################  
 ################################################  
   
 my $have_read_part_table = 0;  my $have_read_part_table = 0;
 my %ids_by_part;  my %ids_by_part;
Line 766  sub get_part { Line 516  sub get_part {
     return undef; # error      return undef; # error
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &get_symb_id()  
   
 Get the MySQL id of a symb.  
   
 Input: $symb  
   
 Output: undef on error, integer $symb_id on success.  
   
 =item &get_symb()  
   
 Get the symb associated with a MySQL symb_id.  
   
 Input: $symb_id  
   
 Output: undef on error, $symb on success.  
   
 =cut  
   
 ################################################  
 ################################################  
   
 my $have_read_symb_table = 0;  my $have_read_symb_table = 0;
 my %ids_by_symb;  my %ids_by_symb;
Line 831  sub get_symb { Line 556  sub get_symb {
     return undef; # error      return undef; # error
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &get_student_id()  
   
 Get the MySQL id of a student.  
   
 Input: $sname, $dom  
   
 Output: undef on error, integer $student_id on success.  
   
 =item &get_student()  
   
 Get student username:domain associated with the MySQL student_id.  
   
 Input: $student_id  
   
 Output: undef on error, string $student (username:domain) on success.  
   
 =cut  
   
 ################################################  
 ################################################  
   
 my $have_read_student_table = 0;  my $have_read_student_table = 0;
 my %ids_by_student;  my %ids_by_student;
 my %students_by_id;  my %students_by_id;
Line 930  sub populate_student_table { Line 629  sub populate_student_table {
     $dbh->do($request);      $dbh->do($request);
     if ($dbh->err()) {      if ($dbh->err()) {
         &Apache::lonnet::logthis("error ".$dbh->errstr().          &Apache::lonnet::logthis("error ".$dbh->errstr().
                                  " occured executing \n".                                   " occurred executing \n".
                                  $request);                                   $request);
     }      }
     return;      return;
Line 983  sub populate_groupnames_table { Line 682  sub populate_groupnames_table {
     $dbh->do($request);      $dbh->do($request);
     if ($dbh->err()) {      if ($dbh->err()) {
         &Apache::lonnet::logthis("error ".$dbh->errstr().          &Apache::lonnet::logthis("error ".$dbh->errstr().
                                  " occured executing \n".                                   " occurred executing \n".
                                  $request);                                   $request);
     }      }
     return;      return;
Line 1053  sub populate_students_groups_table { Line 752  sub populate_students_groups_table {
     $dbh->do($request);      $dbh->do($request);
     if ($dbh->err()) {      if ($dbh->err()) {
         &Apache::lonnet::logthis("error ".$dbh->errstr().          &Apache::lonnet::logthis("error ".$dbh->errstr().
                                  " occured executing \n".                                   " occurred executing \n".
                                  $request);                                   $request);
     }      }
     return;      return;
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &clear_internal_caches()  
   
 Causes the internal caches used in get_student_id, get_student,  
 get_symb_id, get_symb, get_part_id, and get_part to be undef'd.  
   
 Needs to be called before the first operation with the MySQL database  
 for a given Apache request.  
   
 =cut  
   
 ################################################  
 ################################################  
 sub clear_internal_caches {  sub clear_internal_caches {
     $have_read_part_table = 0;      $have_read_part_table = 0;
     undef(%ids_by_part);      undef(%ids_by_part);
Line 1090  sub clear_internal_caches { Line 772  sub clear_internal_caches {
     undef(%ids_by_groupname);      undef(%ids_by_groupname);
 }  }
   
   
 ################################################  
 ################################################  
   
 sub symb_is_for_task {  sub symb_is_for_task {
     my ($symb) = @_;      my ($symb) = @_;
     return ($symb =~ /\.task$/);      return ($symb =~ /\.task$/);
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &update_full_student_data($sname,$sdom,$courseid)  
   
 Does a lonnet::dump on a student to populate the courses tables.  
   
 Input: $sname, $sdom, $courseid  
   
 Output: $returnstatus  
   
 $returnstatus is a string describing any errors that occured.  'okay' is the  
 default.  
   
 This subroutine loads a students data using lonnet::dump and inserts  
 it into the MySQL database.  The inserts are done on three tables,   
 $fulldump_response_table, $fulldump_part_table, and $fulldump_timestamp_table.  
 The INSERT calls are made directly by this subroutine, not through lonmysql   
 because we do a 'bulk'insert which takes advantage of MySQLs non-SQL   
 compliant INSERT command to insert multiple rows at a time.    
 If anything has gone wrong during this process, $returnstatus is updated with   
 a description of the error.  
   
 Once the "fulldump" tables are updated, the tables used for chart and  
 spreadsheet (which hold only the current state of the student on their  
 homework, not historical data) are updated.  If all updates have occured   
 successfully, $student_table is updated to reflect the time of the update.  
   
 Notice we do not insert the data and immediately query it.  This means it  
 is possible for there to be data returned this first time that is not   
 available the second time.  CYA.  
   
 =cut  
   
 ################################################  
 ################################################  
 sub update_full_student_data {  sub update_full_student_data {
     my ($sname,$sdom,$courseid) = @_;      my ($sname,$sdom,$courseid) = @_;
     #      #
Line 1387  sub update_full_student_data { Line 1028  sub update_full_student_data {
     return $returnstatus;      return $returnstatus;
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &update_student_data()  
   
 Input: $sname, $sdom, $courseid  
   
 Output: $returnstatus, \%student_data  
   
 $returnstatus is a string describing any errors that occured.  'okay' is the  
 default.  
 \%student_data is the data returned by a call to lonnet::currentdump.  
   
 This subroutine loads a students data using lonnet::currentdump and inserts  
 it into the MySQL database.  The inserts are done on two tables,   
 $performance_table and $parameters_table.  $parameters_table holds the data   
 that is not included in $performance_table.  See the description of   
 $performance_table elsewhere in this file.  The INSERT calls are made  
 directly by this subroutine, not through lonmysql because we do a 'bulk'  
 insert which takes advantage of MySQLs non-SQL compliant INSERT command to   
 insert multiple rows at a time.  If anything has gone wrong during this  
 process, $returnstatus is updated with a description of the error and  
 \%student_data is returned.    
   
 Notice we do not insert the data and immediately query it.  This means it  
 is possible for there to be data returned this first time that is not   
 available the second time.  CYA.  
   
 =cut  
   
 ################################################  
 ################################################  
 sub update_student_data {  sub update_student_data {
     my ($sname,$sdom,$courseid) = @_;      my ($sname,$sdom,$courseid) = @_;
     #      #
Line 1587  sub store_student_data { Line 1195  sub store_student_data {
     return ($returnstatus,$student_data);      return ($returnstatus,$student_data);
 }  }
   
 ######################################  
 ######################################  
   
 =pod  
   
 =item &ensure_tables_are_set_up($courseid)  
   
 Checks to be sure the MySQL tables for the given class are set up.  
 If $courseid is omitted it will be obtained from the environment.  
   
 Returns nothing on success and 'error' on failure  
   
 =cut  
   
 ######################################  
 ######################################  
 sub ensure_tables_are_set_up {  sub ensure_tables_are_set_up {
     my ($courseid) = @_;      my ($courseid) = @_;
     $courseid = $env{'request.course.id'} if (! defined($courseid));      $courseid = $env{'request.course.id'} if (! defined($courseid));
Line 1640  sub ensure_tables_are_set_up { Line 1233  sub ensure_tables_are_set_up {
     }      }
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &ensure_current_data()  
   
 Input: $sname, $sdom, $courseid  
   
 Output: $status, $data  
   
 This routine ensures the data for a given student is up to date.  
 The $student_table is queried to determine the time of the last update.    
 If the students data is out of date, &update_student_data() is called.    
 The return values from the call to &update_student_data() are returned.  
   
 =cut  
   
 ################################################  
 ################################################  
 sub ensure_current_data {  sub ensure_current_data {
     my ($sname,$sdom,$courseid) = @_;      my ($sname,$sdom,$courseid) = @_;
     my $status = 'okay';   # return value      my $status = 'okay';   # return value
Line 1669  sub ensure_current_data { Line 1242  sub ensure_current_data {
     #      #
     # Get the update time for the user      # Get the update time for the user
     my $updatetime = 0;      my $updatetime = 0;
       my $getuserdir = 1;
     my $modifiedtime = &Apache::lonnet::GetFileTimestamp      my $modifiedtime = &Apache::lonnet::GetFileTimestamp
         ($sdom,$sname,$courseid.'.db',          ($sdom,$sname,$courseid.'.db',$getuserdir);
          $Apache::lonnet::perlvar{'lonUsersDir'});  
     #      #
     if ($modifiedtime == -1) {      if ($modifiedtime == -1) {
  return ('no data',undef);   return ('no data',undef);
Line 1690  sub ensure_current_data { Line 1263  sub ensure_current_data {
     return ($status,$data);      return ($status,$data);
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &ensure_current_full_data($sname,$sdom,$courseid)  
   
 Input: $sname, $sdom, $courseid  
   
 Output: $status  
   
 This routine ensures the fulldata (the data from a lonnet::dump, not a  
 lonnet::currentdump) for a given student is up to date.  
 The $student_table is queried to determine the time of the last update.    
 If the students fulldata is out of date, &update_full_student_data() is  
 called.    
   
 The return value from the call to &update_full_student_data() is returned.  
   
 =cut  
   
 ################################################  
 ################################################  
 sub ensure_current_full_data {  sub ensure_current_full_data {
     my ($sname,$sdom,$courseid) = @_;      my ($sname,$sdom,$courseid) = @_;
     my $status = 'okay';   # return value      my $status = 'okay';   # return value
Line 1721  sub ensure_current_full_data { Line 1272  sub ensure_current_full_data {
     &ensure_tables_are_set_up($courseid);      &ensure_tables_are_set_up($courseid);
     #      #
     # Get the update time for the user      # Get the update time for the user
       my $getuserdir = 1;
     my $modifiedtime = &Apache::lonnet::GetFileTimestamp      my $modifiedtime = &Apache::lonnet::GetFileTimestamp
         ($sdom,$sname,$courseid.'.db',          ($sdom,$sname,$courseid.'.db',$getuserdir);
          $Apache::lonnet::perlvar{'lonUsersDir'});  
     #      #
     my $student_id = &get_student_id($sname,$sdom);      my $student_id = &get_student_id($sname,$sdom);
     my @Result = &Apache::lonmysql::get_rows($student_table,      my @Result = &Apache::lonmysql::get_rows($student_table,
Line 1738  sub ensure_current_full_data { Line 1289  sub ensure_current_full_data {
     return $status;      return $status;
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &get_student_data_from_performance_cache()  
   
 Input: $sname, $sdom, $symb, $courseid  
   
 Output: hash reference containing the data for the given student.  
 If $symb is undef, all the students data is returned.  
   
 This routine is the heart of the local caching system.  See the description  
 of $performance_table, $symb_table, $student_table, and $part_table.  The  
 main task is building the MySQL request.  The tables appear in the request  
 in the order in which they should be parsed by MySQL.  When searching  
 on a student the $student_table is used to locate the 'student_id'.  All  
 rows in $performance_table which have a matching 'student_id' are returned,  
 with data from $part_table and $symb_table which match the entries in  
 $performance_table, 'part_id' and 'symb_id'.  When searching on a symb,  
 the $symb_table is processed first, with matching rows grabbed from   
 $performance_table and filled in from $part_table and $student_table in  
 that order.    
   
 Running 'EXPLAIN ' on the 'SELECT' statements generated can be quite   
 interesting, especially if you play with the order the tables are listed.    
   
 =cut  
   
 ################################################  
 ################################################  
 sub get_student_data_from_performance_cache {  sub get_student_data_from_performance_cache {
     my ($sname,$sdom,$symb,$courseid)=@_;      my ($sname,$sdom,$symb,$courseid)=@_;
     my $student = $sname.':'.$sdom if (defined($sname) && defined($sdom));      my $student = $sname.':'.$sdom if (defined($sname) && defined($sdom));
Line 1852  sub get_student_data_from_performance_ca Line 1373  sub get_student_data_from_performance_ca
     return $studentdata;      return $studentdata;
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &get_current_state()  
   
 Input: $sname,$sdom,$symb,$courseid  
   
 Output: Described below  
   
 Retrieve the current status of a students performance.  $sname and  
 $sdom are the only required parameters.  If $symb is undef the results  
 of an &Apache::lonnet::currentdump() will be returned.    
 If $courseid is undef it will be retrieved from the environment.  
   
 The return structure is based on &Apache::lonnet::currentdump.  If  
 $symb is unspecified, all the students data is returned in a hash of  
 the form:  
 (   
   symb1 => { param1 => value1, param2 => value2 ... },  
   symb2 => { param1 => value1, param2 => value2 ... },  
 )  
   
 If $symb is specified, a hash of   
 (  
   param1 => value1,   
   param2 => value2,  
 )  
 is returned.  
   
 If no data is found for $symb, or if the student has no performance data,  
 an empty list is returned.  
   
 =cut  
   
 ################################################  
 ################################################  
 sub get_current_state {  sub get_current_state {
     my ($sname,$sdom,$symb,$courseid,$forcedownload)=@_;      my ($sname,$sdom,$symb,$courseid,$forcedownload)=@_;
     #      #
Line 1929  sub get_current_state { Line 1413  sub get_current_state {
     return ();      return ();
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &get_problem_statistics()  
   
 Gather data on a given problem.  The database is assumed to be   
 populated and all local caching variables are assumed to be set  
 properly.  This means you need to call &ensure_current_data for  
 the students you are concerned with prior to calling this routine.  
   
 Inputs: $Sections, Groups, $status, $symb, $part, $courseid, $starttime,  
         $endtime  
   
 =over 4  
   
 =item $Sections Array ref containing section names for students.    
 'all' is allowed to be the first (and only) item in the array.  
   
 =item $Groups Array ref containing group names for students.  
 'all' is allowed to be the first (and only) item in the array.  
   
 =item $status String describing the status of students  
   
 =item $symb is the symb for the problem.  
   
 =item $part is the part id you need statistics for  
   
 =item $courseid is the course id, of course!  
   
 =item $starttime and $endtime are unix times which to use to limit  
 the statistical data.  
   
 =back  
   
 Outputs: See the code for up to date information.  A hash reference is  
 returned.  The hash has the following keys defined:  
   
 =over 4  
   
 =item num_students The number of students attempting the problem  
         
 =item tries The total number of tries for the students  
         
 =item max_tries The maximum number of tries taken  
         
 =item mean_tries The average number of tries  
         
 =item num_solved The number of students able to solve the problem  
         
 =item num_override The number of students whose answer is 'correct_by_override'  
         
 =item deg_of_diff The degree of difficulty of the problem  
         
 =item std_tries The standard deviation of the number of tries  
         
 =item skew_tries The skew of the number of tries  
   
 =item per_wrong The number of students attempting the problem who were not  
 able to answer it correctly.  
   
 =back  
   
 =cut  
   
 ################################################  
 ################################################  
 sub get_problem_statistics {  sub get_problem_statistics {
     my ($Sections,$Groups,$status,$symb,$part,$courseid,$starttime,$endtime) = @_;      my ($Sections,$Groups,$status,$symb,$part,$courseid,$starttime,$endtime) = @_;
     return if (! defined($symb) || ! defined($part));      return if (! defined($symb) || ! defined($part));
Line 2146  sub execute_SQL_request { Line 1563  sub execute_SQL_request {
     return ();      return ();
 }  }
   
 ######################################################  
 ######################################################  
   
 =pod  
   
 =item &populate_weight_table  
   
 =cut  
   
 ######################################################  
 ######################################################  
 sub populate_weight_table {  sub populate_weight_table {
     my ($courseid) = @_;      my ($courseid) = @_;
     if (! defined($courseid)) {      if (! defined($courseid)) {
Line 2209  sub populate_weight_table { Line 1616  sub populate_weight_table {
     $dbh->do($request);      $dbh->do($request);
     if ($dbh->err()) {      if ($dbh->err()) {
         &Apache::lonnet::logthis("error ".$dbh->errstr().          &Apache::lonnet::logthis("error ".$dbh->errstr().
                                  " occured executing \n".                                   " occurred executing \n".
                                  $request);                                   $request);
     }      }
     return;      return;
 }  }
   
 ##########################################################  
 ##########################################################  
   
 =pod  
   
 =item &limit_by_start_end_times  
   
 Build SQL WHERE condition which limits the data collected by the start  
 and end times provided  
   
 Inputs: $starttime, $endtime, $table  
   
 Returns: $time_limits  
   
 =cut  
   
 ##########################################################  
 ##########################################################  
 sub limit_by_start_end_time {  sub limit_by_start_end_time {
     my ($starttime,$endtime,$table) = @_;      my ($starttime,$endtime,$table) = @_;
     my $time_requirements = undef;      my $time_requirements = undef;
Line 2247  sub limit_by_start_end_time { Line 1636  sub limit_by_start_end_time {
     return $time_requirements;      return $time_requirements;
 }  }
   
 ##########################################################  
 ##########################################################  
   
 =pod  
   
 =item &limit_by_section_and_status  
   
 Build SQL WHERE condition which limits the data collected by section and  
 student status.  
   
 Inputs: $Sections (array ref)  
     $enrollment (string: 'any', 'expired', 'active')  
     $tablename The name of the table that holds the student data  
   
 Returns: $student_requirements,$enrollment_requirements  
   
 =cut  
   
 ##########################################################  
 ##########################################################  
 sub limit_by_section_and_status {  sub limit_by_section_and_status {
     my ($Sections,$enrollment,$tablename) = @_;      my ($Sections,$enrollment,$tablename) = @_;
     my $student_requirements = undef;      my $student_requirements = undef;
Line 2295  sub limit_by_section_and_status { Line 1665  sub limit_by_section_and_status {
     return ($student_requirements,$enrollment_requirements);      return ($student_requirements,$enrollment_requirements);
 }  }
   
 ######################################################  
 ######################################################  
                                                                                  
 =pod  
                                                                                  
 =item &limit_by_group  
                                                                                  
 Build SQL LEFT JOIN statement to include students_groups and groupnames tables and SQL WHERE condition which limits the data collected by group.  
                                                                                  
 Inputs: $Groups (array ref)  
     $stutable   The name of the table which holds the student data.  
     $grptable   The name of the table which maps group_id to groupname.  
     $stugrptab  The name of the table which holds student group affiliations.     
 Returns: $groups_join,$group_limits  
    $groups_join  JOIN part of SQL statement (to include group related tables)   
    $group_limits SQL WHERE condition limiting to requested groups  
 =cut  
   
 sub limit_by_group {  sub limit_by_group {
     my ($Groups,$stutable,$grptable,$stugrptab) = @_;      my ($Groups,$stutable,$grptable,$stugrptab) = @_;
Line 2333  sub limit_by_group { Line 1687  sub limit_by_group {
     return ($groups_join,$group_limits);      return ($groups_join,$group_limits);
 }  }
   
 =pod  
   
 =item rank_students_by_scores_on_resources  
   
 Inputs:   
     $resources: array ref of hash ref.  Each hash ref needs key 'symb'.  
     $Sections: array ref of sections to include,  
     $Groups: array ref of groups to include.  
     $enrollment: string,  
     $courseid (may be omitted)  
     $starttime (may be omitted)  
     $endtime (may be omitted)  
     $has_award_for (may be omitted)  
   
 Returns; An array of arrays.  The sub arrays contain a student name and  
 their score on the resources. $starttime and $endtime constrain the  
 list to awards obtained during the given time limits. $has_score_on  
 constrains the list to those students who at least attempted the  
 resource identified by the given symb, which is used to filter out  
 such students for statistics that would be adversely affected by such  
 students.   
   
 =cut  
   
 ######################################################  
 ######################################################  
 sub RNK_student { return 0; };  sub RNK_student { return 0; };
 sub RNK_score   { return 1; };  sub RNK_score   { return 1; };
   
Line 2420  sub rank_students_by_scores_on_resources Line 1749  sub rank_students_by_scores_on_resources
     return ($rows);      return ($rows);
 }  }
   
 ########################################################  
 ########################################################  
   
 =pod  
   
 =item &get_sum_of_scores  
   
 Inputs: $resource (hash ref, needs {'symb'} key),  
 $part, (the part id),  
 $students (array ref, contents of array are scalars holding 'sname:sdom'),  
 $courseid  
   
 Returns: the sum of the score on the problem part over the students and the  
    maximum possible value for the sum (taken from the weight table).  
   
 =cut  
   
 ########################################################  
 ########################################################  
 sub get_sum_of_scores {  sub get_sum_of_scores {
     my ($symb,$part,$students,$courseid,$starttime,$endtime) = @_;      my ($symb,$part,$students,$courseid,$starttime,$endtime) = @_;
     if (! defined($courseid)) {      if (! defined($courseid)) {
Line 2482  sub get_sum_of_scores { Line 1792  sub get_sum_of_scores {
     return ($rows->[0],$rows->[1]);      return ($rows->[0],$rows->[1]);
 }  }
   
 ########################################################  
 ########################################################  
   
 =pod  
   
 =item &score_stats  
   
 Inputs: $Sections, $enrollment, $symbs, $starttime,  
         $endtime, $courseid  
   
 $Sections, $enrollment, $starttime, $endtime, and $courseid are the same as   
 elsewhere in this module.    
 $symbs is an array ref of symbs  
   
 Returns: minimum, maximum, mean, s.d., number of students, and maximum  
   possible of student scores on the given resources  
   
 =cut  
   
 ########################################################  
 ########################################################  
 sub score_stats {  sub score_stats {
     my ($Sections,$Groups,$enrollment,$symbs,$starttime,$endtime,$courseid)=@_;      my ($Sections,$Groups,$enrollment,$symbs,$starttime,$endtime,$courseid)=@_;
     if (! defined($courseid)) {      if (! defined($courseid)) {
Line 2561  sub score_stats { Line 1851  sub score_stats {
 }  }
   
   
 ########################################################  
 ########################################################  
   
 =pod  
   
 =item &count_stats  
   
 Inputs: $Sections, $Groups, $enrollment, $symbs, $starttime,  
         $endtime, $courseid  
   
 $Sections, $Groups $enrollment, $starttime, $endtime, and $courseid are the   
 same as elsewhere in this module.    
 $symbs is an array ref of symbs  
   
 Returns: minimum, maximum, mean, s.d., and number of students  
   of the number of items correct on the given resources  
   
 =cut  
   
 ########################################################  
 ########################################################  
 sub count_stats {  sub count_stats {
     my ($Sections,$Groups,$enrollment,$symbs,$starttime,$endtime,$courseid)=@_;      my ($Sections,$Groups,$enrollment,$symbs,$starttime,$endtime,$courseid)=@_;
     if (! defined($courseid)) {      if (! defined($courseid)) {
Line 2633  sub count_stats { Line 1903  sub count_stats {
     return($min,$max,$ave,$std,$count);      return($min,$max,$ave,$std,$count);
 }  }
   
 ######################################################  
 ######################################################  
   
 =pod  
   
 =item get_student_data  
   
 =cut  
   
 ######################################################  
 ######################################################  
 sub get_student_data {  sub get_student_data {
     my ($students,$courseid) = @_;      my ($students,$courseid) = @_;
     $courseid = $env{'request.course.id'} if (! defined($courseid));      $courseid = $env{'request.course.id'} if (! defined($courseid));
Line 2878  sub get_response_time_data { Line 2138  sub get_response_time_data {
   
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &get_student_scores($Sections,$Groups,$Symbs,$enrollment,$courseid)  
   
 =cut  
   
 ################################################  
 ################################################  
 sub get_student_scores {  sub get_student_scores {
     my ($sections,$groups,$Symbs,$enrollment,$courseid,$starttime,$endtime) = @_;      my ($sections,$groups,$Symbs,$enrollment,$courseid,$starttime,$endtime) = @_;
     $courseid = $env{'request.course.id'} if (! defined($courseid));      $courseid = $env{'request.course.id'} if (! defined($courseid));
Line 2966  sub get_student_scores { Line 2215  sub get_student_scores {
     return $dataset;      return $dataset;
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =item &setup_table_names()  
   
 input: course id  
   
 output: none  
   
 Cleans up the package variables for local caching.  
   
 =cut  
   
 ################################################  
 ################################################  
 sub setup_table_names {  sub setup_table_names {
     my ($courseid) = @_;      my ($courseid) = @_;
     if (! defined($courseid)) {      if (! defined($courseid)) {
Line 3042  sub temp_table_name { Line 2276  sub temp_table_name {
     return $base_id.'_'.$affix;      return $base_id.'_'.$affix;
 }  }
   
 ################################################  
 ################################################  
   
 =pod  
   
 =back  
   
 =item End of Local Data Caching Subroutines  
   
 =cut  
   
 ################################################  
 ################################################  
   
 } # End scope of table identifiers  } # End scope of table identifiers
   
 ################################################  
 ################################################  
   
 =pod  
   
 =head3 Classlist Subroutines  
   
 =item &get_classlist();  
   
 Retrieve the classist of a given class or of the current class.  Student  
 information is returned from the classlist.db file and, if needed,  
 from the students environment.  
   
 Optional arguments are $cdom, and $cnum (course domain,  
 and course number, respectively).  If either is ommitted the course  
 will be taken from the current environment ($env{'request.course.id'},  
 $env{'course.'.$cid.'.domain'}, and $env{'course.'.$cid.'.num'}).  
   
 Returns a reference to a hash which contains:  
  keys    '$sname:$sdom'  
  values  [$sdom,$sname,$end,$start,$id,$section,$fullname,$status,$type,$lockedtype]  
   
 The constant values CL_SDOM, CL_SNAME, CL_END, etc. can be used  
 as indices into the returned list to future-proof clients against  
 changes in the list order.  
   
 =cut  
   
 ################################################  
 ################################################  
   
 sub CL_SDOM     { return 0; }  sub CL_SDOM     { return 0; }
 sub CL_SNAME    { return 1; }  sub CL_SNAME    { return 1; }
Line 3099  sub CL_FULLNAME { return 6; } Line 2291  sub CL_FULLNAME { return 6; }
 sub CL_STATUS   { return 7; }  sub CL_STATUS   { return 7; }
 sub CL_TYPE     { return 8; }  sub CL_TYPE     { return 8; }
 sub CL_LOCKEDTYPE   { return 9; }  sub CL_LOCKEDTYPE   { return 9; }
 sub CL_GROUP    { return 10; }  sub CL_CREDITS  { return 10; }
   sub CL_GROUP    { return 11; }
   sub CL_PERMANENTEMAIL { return 12; }
   sub CL_ROLE     { return 13; }
   sub CL_EXTENT   { return 14; }
   sub CL_PHOTO   { return 15; }
   sub CL_THUMBNAIL { return 16; }
   sub CL_AUTHORQUOTA { return 17; }
   sub CL_AUTHORUSAGE { return 18; }
   
 sub get_classlist {  sub get_classlist {
     my ($cdom,$cnum) = @_;      my ($cdom,$cnum) = @_;
Line 3119  sub get_classlist { Line 2319  sub get_classlist {
         }          }
         my ($sname,$sdom) = split(/:/,$student);          my ($sname,$sdom) = split(/:/,$student);
         my @Values = split(/:/,$info);          my @Values = split(/:/,$info);
         my ($end,$start,$id,$section,$fullname,$type,$lockedtype);          my ($end,$start,$id,$section,$fullname,$type,$lockedtype,$credits);
         if (@Values > 2) {          if (@Values > 2) {
             ($end,$start,$id,$section,$fullname,$type,$lockedtype) = @Values;              ($end,$start,$id,$section,$fullname,$type,$lockedtype,$credits) = @Values;
         } else { # We have to get the data ourselves          } else { # We have to get the data ourselves
             ($end,$start) = @Values;              ($end,$start) = @Values;
             $section = &Apache::lonnet::getsection($sdom,$sname,$cid);              $section = &Apache::lonnet::getsection($sdom,$sname,$cid);
Line 3160  sub get_classlist { Line 2360  sub get_classlist {
             $status='Future';              $status='Future';
         }          }
         $classlist{$student} =           $classlist{$student} = 
             [$sdom,$sname,$end,$start,$id,$section,$fullname,$status,$type,$lockedtype];              [$sdom,$sname,$end,$start,$id,$section,$fullname,$status,$type,
                $lockedtype,$credits];
     }      }
     if (wantarray()) {      if (wantarray()) {
         return (\%classlist,['domain','username','end','start','id',          return (\%classlist,['domain','username','end','start','id',
                              'section','fullname','status','type','lockedtype']);                               'section','fullname','status','type',
                                'lockedtype','credits']);
     } else {      } else {
         return \%classlist;          return \%classlist;
     }      }
Line 3263  sub get_students_groups { Line 2465  sub get_students_groups {
 1;  1;
 __END__  __END__
   
   
   =pod
   
   =head1 NAME
   
   Apache::loncoursedata
   
   =head1 SYNOPSIS
   
   Set of functions that download and process student and course information.
   
   =head1 PACKAGES USED
   
     Apache::lonnet
     Apache::longroup
     Time::HiRes
     Apache::lonmysql
     LONCAPA
     Digest::MD5
    
   =head1 DOWNLOAD INFORMATION
   
   This section contains all the functions that get data from other servers 
   and/or itself.
   
   
   
   =head1 LOCAL DATA CACHING SUBROUTINES
   
   The local caching is done using MySQL.  There is no fall-back implementation
   if MySQL is not running.
   
   The programmers interface is to call &get_current_state() or some other
   primary interface subroutine (described below).  The internals of this 
   storage system are documented here.
   
   There are six tables used to store student performance data (the results of
   a dumpcurrent).  Each of these tables is created in MySQL with a name of
   $courseid_*****, where ***** is 'symb', 'part', or whatever is appropriate 
   for the table.  The tables and their purposes are described below.
   
   Some notes before we get started.
   
   Each table must have a PRIMARY KEY, which is a column or set of columns which
   will serve to uniquely identify a row of data.  NULL is not allowed!
   
   INDEXes work best on integer data.
   
   JOIN is used to combine data from many tables into one output.
   
   lonmysql.pm is used for some of the interface, specifically the table creation
   calls.  The inserts are done in bulk by directly calling the database handler.
   The SELECT ... JOIN statement used to retrieve the data does not have an
   interface in lonmysql.pm and I shudder at the thought of writing one.
   
   =head2 Table Descriptions
   
   =over 4
   
   =head2 Tables used to store meta information
   
   The following tables hold data required to keep track of the current status
   of a students data in the tables or to look up the students data in the tables.
   
   =over 4
   
   =item C<$symb_table>
   
   The symb_table has two columns.  The first is a 'symb_id' and the second
   is the text name for the 'symb' (limited to 64k).  The 'symb_id' is generated
   automatically by MySQL so inserts should be done on this table with an
   empty first element.  This table has its PRIMARY KEY on the 'symb_id'.
   
   =item C<$part_table>
   
   The part_table has two columns.  The first is a 'part_id' and the second
   is the text name for the 'part' (limited to 100 characters).  The 'part_id' is
   generated automatically by MySQL so inserts should be done on this table with
   an empty first element.  This table has its PRIMARY KEY on the 'part' (100
   characters) and a KEY on 'part_id'.
   
   =item C<$student_table>
   
   The student_table has 7 columns.  The first is a 'student_id' assigned by 
   MySQL.  The second is 'student' which is username:domain.  The third through
   fifth are 'section', 'status' (enrollment status), and 'classification' 
   (to be used in the future).  The sixth and seventh ('updatetime' and 
   'fullupdatetime') contain the time of last update and full update of student
   data.  This table has its PRIMARY KEY on the 'student_id' column and is indexed
   on 'student', 'section', and 'status'.
   
   =item C<$groupnames_table>
   
   The groupnames_table has 2 columns.  The first is a 'group_id' assigned by 
   MySQL.  The second is 'groupname' which is the name of the group in the course.
   
   =item C<$students_groups_table>
   
   The students_groups_table has 2 columns.  The first is the 'student_id', and the 
   second is the 'group_id'. These two columns comprise the PRIMARY KEY for this 
   table, as an individual student may be affiliated with more than one group at
   any time. This table is indexed on both student_id and group_id.
   
   =back 
   
   =head2 Tables used to store current status data
   
   The following tables store data only about the students current status on 
   a problem, meaning only the data related to the last attempt on a problem.
   
   =over 4
   
   =item C<$performance_table>
   
   The performance_table has 9 columns.  The first three are 'symb_id', 
   'student_id', and 'part_id'.  These comprise the PRIMARY KEY for this table
   and are directly related to the $symb_table, $student_table, and $part_table
   described above.  MySQL does better indexing on numeric items than text,
   so we use these three "index tables".  The remaining columns are
   'solved', 'tries', 'awarded', 'award', 'awarddetail', and 'timestamp'.
   These are either the MySQL type TINYTEXT or various integers ('tries' and 
   'timestamp').  This table has KEYs of 'student_id' and 'symb_id'.
   For use of this table, see the functions described below.
   
   =item C<$parameters_table>
   
   The parameters_table holds the data that does not fit neatly into the
   performance_table.  The parameters table has four columns: 'symb_id',
   'student_id', 'parameter', and 'value'.  'symb_id', 'student_id', and
   'parameter' comprise the PRIMARY KEY for this table.  'parameter' is 
   limited to 255 characters.  'value' is limited to 64k characters.
   
   =back
   
   =head2 Tables used for storing historic data
   
   The following tables are used to store almost all of the transactions a student
   has made on a homework problem.  See loncapa/docs/homework/datastorage for 
   specific information about each of the parameters stored.  
   
   =over 4
   
   =item C<$fulldump_response_table>
   
   The response table holds data (documented in loncapa/docs/homework/datastorage)
   associated with a particular response id which is stored when a student 
   attempts a problem.  The following are the columns of the table, in order:
   'symb_id','part_id','response_id','student_id','transaction','tries',
   'awarddetail', 'response_specific', 'response_specific_value',
   'response_specific_2', 'response_specific_value_2', and 'submission
   (the text of the students submission).  The primary key is based on the
   first five columns listed above.
   
   =item C<$fulldump_part_table()>
   
   The part table holds data (documented in loncapa/docs/homework/datastorage)
   associated with a particular part id which is stored when a student attempts
   a problem.  The following are the columns of the table, in order:
   'symb_id','part_id','student_id','transaction','tries','award','awarded',
   and 'previous'.  The primary key is based on the first five columns listed 
   above.
   
   =item C<$fulldump_timestamp_table()>
   
   The timestamp table holds the timestamps of the transactions which are
   stored in $fulldump_response_table and $fulldump_part_table.  This data is
   about both the response and part data.  Columns: 'symb_id','student_id',
   'transaction', and 'timestamp'.  
   The primary key is based on the first 3 columns.
   
   =item C<$weight_table()>
   
   The weight table holds the weight for the problems used in the class.
   Whereas the weight of a problem can vary by section and student the data
   here is applied to the class as a whole.
   Columns: 'symb_id','part_id','response_id','weight'.
   
   =back
   
   
   =head1 IMPORTANT SUBROUTINES
   
   Here is a brief overview of the subroutines which are likely to be of 
   interest:
   
   =over 4
   
   =item C<&get_current_state()>
   
   programmers interface.
   
   =item C<&init_dbs()>
   
   table creation
   
   =item C<&update_student_data()>
   
   data storage calls
   
   =item C<&get_student_data_from_performance_cache()>
   
   data retrieval
   
   =back
   
   =head1 OTHER SUBROUTINES
   
   =over 4
   
   =item C<&make_into_hash($values)>
   
   Returns a reference to a hash as described by $values.  $values is
   assumed to be the result of 
       join(':',map {&escape($_)} %orighash);
   
   This is a helper function for get_current_state.
   
   =item C<&init_dbs()>
   
   Input: course id
   
   Output: 0 on success, positive integer on error
   
   This routine issues the calls to lonmysql to create the tables used to
   store student data.
   
   item C<&delete_caches()>
   
   This routine drops all the tables associated with a course from the 
   MySQL database.
   
   Input: course id (optional, determined by environment if omitted) 
   
   Returns: nothing
   
   =item C<&get_part_id()>
   
   Get the MySQL id of a problem part string.
   
   Input: $part
   
   Output: undef on error, integer $part_id on success.
   
   =item C<&get_part()>
   
   Get the string describing a part from the MySQL id of the problem part.
   
   Input: $part_id
   
   Output: undef on error, $part string on success.
   
   =item C<&get_symb_id()>
   
   Get the MySQL id of a symb.
   
   Input: $symb
   
   Output: undef on error, integer $symb_id on success.
   
   =item C<&get_symb()>
   
   Get the symb associated with a MySQL symb_id.
   
   Input: $symb_id
   
   Output: undef on error, $symb on success.
   
   =item C<&get_student_id()>
   
   Get the MySQL id of a student.
   
   Input: $sname, $dom
   
   Output: undef on error, integer $student_id on success.
   
   =item C<&get_student()>
   
   Get student username:domain associated with the MySQL student_id.
   
   Input: $student_id
   
   Output: undef on error, string $student (username:domain) on success.
   
   =item C<&clear_internal_caches()>
   
   Causes the internal caches used in get_student_id, get_student,
   get_symb_id, get_symb, get_part_id, and get_part to be undef'd.
   
   Needs to be called before the first operation with the MySQL database
   for a given Apache request.
   
   =item C<&update_full_student_data($sname,$sdom,$courseid)>
   
   Does a lonnet::dump on a student to populate the courses tables.
   
   Input: $sname, $sdom, $courseid
   
   Output: $returnstatus
   
   $returnstatus is a string describing any errors that occurred.  'okay' is the
   default.
   
   This subroutine loads a students data using lonnet::dump and inserts
   it into the MySQL database.  The inserts are done on three tables, 
   $fulldump_response_table, $fulldump_part_table, and $fulldump_timestamp_table.
   The INSERT calls are made directly by this subroutine, not through lonmysql 
   because we do a 'bulk'insert which takes advantage of MySQLs non-SQL 
   compliant INSERT command to insert multiple rows at a time.  
   If anything has gone wrong during this process, $returnstatus is updated with 
   a description of the error.
   
   Once the "fulldump" tables are updated, the tables used for chart and
   spreadsheet (which hold only the current state of the student on their
   homework, not historical data) are updated.  If all updates have occurred 
   successfully, $student_table is updated to reflect the time of the update.
   
   Notice we do not insert the data and immediately query it.  This means it
   is possible for there to be data returned this first time that is not 
   available the second time.  CYA.
   
   
   =item C<&update_student_data()>
   
   Input: $sname, $sdom, $courseid
   
   Output: $returnstatus, \%student_data
   
   $returnstatus is a string describing any errors that occurred.  'okay' is the
   default.
   \%student_data is the data returned by a call to lonnet::currentdump.
   
   This subroutine loads a students data using lonnet::currentdump and inserts
   it into the MySQL database.  The inserts are done on two tables, 
   $performance_table and $parameters_table.  $parameters_table holds the data 
   that is not included in $performance_table.  See the description of 
   $performance_table elsewhere in this file.  The INSERT calls are made
   directly by this subroutine, not through lonmysql because we do a 'bulk'
   insert which takes advantage of MySQLs non-SQL compliant INSERT command to 
   insert multiple rows at a time.  If anything has gone wrong during this
   process, $returnstatus is updated with a description of the error and
   \%student_data is returned.  
   
   Notice we do not insert the data and immediately query it.  This means it
   is possible for there to be data returned this first time that is not 
   available the second time.  CYA.
   
   =item &ensure_tables_are_set_up($courseid)
   
   Checks to be sure the MySQL tables for the given class are set up.
   If $courseid is omitted it will be obtained from the environment.
   
   Returns nothing on success and 'error' on failure
   
   
   =item C<&ensure_current_data()>
   
   Input: $sname, $sdom, $courseid
   
   Output: $status, $data
   
   This routine ensures the data for a given student is up to date.
   The $student_table is queried to determine the time of the last update.  
   If the students data is out of date, &update_student_data() is called.  
   The return values from the call to &update_student_data() are returned.
   
   =item C<&ensure_current_full_data($sname,$sdom,$courseid)>
   
   Input: $sname, $sdom, $courseid
   
   Output: $status
   
   This routine ensures the fulldata (the data from a lonnet::dump, not a
   lonnet::currentdump) for a given student is up to date.
   The $student_table is queried to determine the time of the last update.  
   If the students fulldata is out of date, &update_full_student_data() is
   called.  
   
   The return value from the call to &update_full_student_data() is returned.
   
   =item C<&get_student_data_from_performance_cache()>
   
   Input: $sname, $sdom, $symb, $courseid
   
   Output: hash reference containing the data for the given student.
   If $symb is undef, all the students data is returned.
   
   This routine is the heart of the local caching system.  See the description
   of $performance_table, $symb_table, $student_table, and $part_table.  The
   main task is building the MySQL request.  The tables appear in the request
   in the order in which they should be parsed by MySQL.  When searching
   on a student the $student_table is used to locate the 'student_id'.  All
   rows in $performance_table which have a matching 'student_id' are returned,
   with data from $part_table and $symb_table which match the entries in
   $performance_table, 'part_id' and 'symb_id'.  When searching on a symb,
   the $symb_table is processed first, with matching rows grabbed from 
   $performance_table and filled in from $part_table and $student_table in
   that order.  
   
   Running 'EXPLAIN ' on the 'SELECT' statements generated can be quite 
   interesting, especially if you play with the order the tables are listed.  
   
   
   =item C<&get_current_state()>
   
   Input: $sname,$sdom,$symb,$courseid
   
   Output: Described below
   
   Retrieve the current status of a students performance.  $sname and
   $sdom are the only required parameters.  If $symb is undef the results
   of an &Apache::lonnet::currentdump() will be returned.  
   If $courseid is undef it will be retrieved from the environment.
   
   The return structure is based on &Apache::lonnet::currentdump.  If
   $symb is unspecified, all the students data is returned in a hash of
   the form:
   ( 
     symb1 => { param1 => value1, param2 => value2 ... },
     symb2 => { param1 => value1, param2 => value2 ... },
   )
   
   If $symb is specified, a hash of 
   (
     param1 => value1, 
     param2 => value2,
   )
   is returned.
   
   If no data is found for $symb, or if the student has no performance data,
   an empty list is returned.
   
   =item C<&get_problem_statistics()>
   
   Gather data on a given problem.  The database is assumed to be 
   populated and all local caching variables are assumed to be set
   properly.  This means you need to call &ensure_current_data for
   the students you are concerned with prior to calling this routine.
   
   Inputs: $Sections, Groups, $status, $symb, $part, $courseid, $starttime,
           $endtime
   
   =over 4
   
   =item $Sections Array ref containing section names for students.  
   'all' is allowed to be the first (and only) item in the array.
   
   =item $Groups Array ref containing group names for students.
   'all' is allowed to be the first (and only) item in the array.
   
   =item $status String describing the status of students
   
   =item $symb is the symb for the problem.
   
   =item $part is the part id you need statistics for
   
   =item $courseid is the course id, of course!
   
   =item $starttime and $endtime are unix times which to use to limit
   the statistical data.
   
   =back
   
   Outputs: See the code for up to date information.  A hash reference is
   returned.  The hash has the following keys defined:
   
   =over 4
   
   =item * num_students 
   
   The number of students attempting the problem
         
   =item tries 
   
   The total number of tries for the students
         
   =item max_tries 
   
   The maximum number of tries taken
         
   =item mean_tries 
   
   The average number of tries
         
   =item num_solved T
   
   he number of students able to solve the problem
         
   =item num_override 
   
   The number of students whose answer is 'correct_by_override'
         
   =item deg_of_diff 
   
   The degree of difficulty of the problem
         
   =item std_tries 
   
   The standard deviation of the number of tries
         
   =item skew_tries 
   
   The skew of the number of tries
   
   =item per_wrong 
   
   The number of students attempting the problem who were not
   able to answer it correctly.
   
   =back
   
   =item C<&populate_weight_table()>
   
   =item C<&limit_by_start_end_times()>
   
   Build SQL WHERE condition which limits the data collected by the start
   and end times provided
   
   Inputs: $starttime, $endtime, $table
   
   Returns: $time_limits
   
   
   =item C<&limit_by_section_and_status()C<
   
   Build SQL WHERE condition which limits the data collected by section and
   student status.
   
   Inputs: $Sections (array ref)
       $enrollment (string: 'any', 'expired', 'active')
       $tablename The name of the table that holds the student data
   
   Returns: $student_requirements,$enrollment_requirements
   
   =item C<&limit_by_group()>
                                                                                  
   Build SQL LEFT JOIN statement to include students_groups and groupnames tables and SQL WHERE condition which limits the data collected by group.
                                                                                  
   Inputs: $Groups (array ref)
       $stutable   The name of the table which holds the student data.
       $grptable   The name of the table which maps group_id to groupname.
       $stugrptab  The name of the table which holds student group affiliations.   
   Returns: $groups_join,$group_limits
      $groups_join  JOIN part of SQL statement (to include group related tables) 
      $group_limits SQL WHERE condition limiting to requested groups
   
   =item C<rank_students_by_scores_on_resources()>
   
   Inputs: 
       $resources: array ref of hash ref.  Each hash ref needs key 'symb'.
       $Sections: array ref of sections to include,
       $Groups: array ref of groups to include.
       $enrollment: string,
       $courseid (may be omitted)
       $starttime (may be omitted)
       $endtime (may be omitted)
       $has_award_for (may be omitted)
   
   Returns; An array of arrays.  The sub arrays contain a student name and
   their score on the resources. $starttime and $endtime constrain the
   list to awards obtained during the given time limits. $has_score_on
   constrains the list to those students who at least attempted the
   resource identified by the given symb, which is used to filter out
   such students for statistics that would be adversely affected by such
   students.
   
   =item C<&get_sum_of_scores>
   
   Inputs: $resource (hash ref, needs {'symb'} key),
   $part, (the part id),
   $students (array ref, contents of array are scalars holding 'sname:sdom'),
   $courseid
   
   Returns: the sum of the score on the problem part over the students and the
      maximum possible value for the sum (taken from the weight table).
    
   
   =item C<&score_stats()>
   
   Inputs: $Sections, $enrollment, $symbs, $starttime,
           $endtime, $courseid
   
   $Sections, $enrollment, $starttime, $endtime, and $courseid are the same as 
   elsewhere in this module.  
   $symbs is an array ref of symbs
   
   Returns: minimum, maximum, mean, s.d., number of students, and maximum
     possible of student scores on the given resources
   
   =item C<&count_stats()>
   
   Inputs: $Sections, $Groups, $enrollment, $symbs, $starttime,
           $endtime, $courseid
   
   $Sections, $Groups $enrollment, $starttime, $endtime, and $courseid are the 
   same as elsewhere in this module.  
   $symbs is an array ref of symbs
   
   Returns: minimum, maximum, mean, s.d., and number of students
     of the number of items correct on the given resources
   
   =item C<get_student_data()>
   
   =item C<&get_student_scores($Sections,$Groups,$Symbs,$enrollment,$courseid)>
   
   =item C<&setup_table_names()>
   
   input: course id
   
   output: none
   
   =back
   
   =head3 End of Local Data Caching Subroutines
   
   =head3 Classlist Subroutines
   
   =over
   
   =item &get_classlist();
   
   Retrieve the classist of a given class or of the current class.  Student
   information is returned from the classlist.db file and, if needed,
   from the students environment.
   
   Optional arguments are $cdom, and $cnum (course domain,
   and course number, respectively).  If either is ommitted the course
   will be taken from the current environment ($env{'request.course.id'},
   $env{'course.'.$cid.'.domain'}, and $env{'course.'.$cid.'.num'}).
   
   Returns a reference to a hash which contains:
    keys    '$sname:$sdom'
    values  [$sdom,$sname,$end,$start,$id,$section,$fullname,$status,$type,
             $lockedtype,$credits]
   
   The constant values CL_SDOM, CL_SNAME, CL_END, etc. can be used
   as indices into the returned list to future-proof clients against
   changes in the list order.
   
   =back
   
   =cut
   
   
Removed from v.1.183  
changed lines
  Added in v.1.195

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