File:  [LON-CAPA] / loncom / LONCAPA.pm
Revision 1.25: download - view: text, annotated - select for diffs
Mon Nov 17 13:22:01 2008 UTC (15 years, 5 months ago) by jms
Branches: MAIN
CVS tags: HEAD
Added & converted POD documentation

    1: # The LearningOnline Network
    2: # Base routines
    3: #
    4: # $Id: LONCAPA.pm,v 1.25 2008/11/17 13:22:01 jms Exp $
    5: #
    6: # Copyright Michigan State University Board of Trustees
    7: #
    8: # This file is part of the LearningOnline Network with CAPA (LON-CAPA).
    9: #
   10: # LON-CAPA is free software; you can redistribute it and/or modify
   11: # it under the terms of the GNU General Public License as published by
   12: # the Free Software Foundation; either version 2 of the License, or
   13: # (at your option) any later version.
   14: #
   15: # LON-CAPA is distributed in the hope that it will be useful,
   16: # but WITHOUT ANY WARRANTY; without even the implied warranty of
   17: # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   18: # GNU General Public License for more details.
   19: #
   20: # You should have received a copy of the GNU General Public License
   21: # along with LON-CAPA; if not, write to the Free Software
   22: # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
   23: #
   24: # /home/httpd/html/adm/gpl.txt
   25: #
   26: # http://www.lon-capa.org/
   27: #
   28: ###
   29: 
   30: =head1 NAME
   31: 
   32: Apache::LONCAPA
   33: 
   34: LONCAPA - Basic routines
   35: 
   36: =head1 SYNOPSIS
   37: 
   38: Generally useful routines
   39: 
   40: =head1 EXPORTED SUBROUTINES
   41: 
   42: =over 4
   43: 
   44: =item *
   45: 
   46: escape() : unpack non-word characters into CGI-compatible hex codes
   47: 
   48: =item *
   49: 
   50: unescape() : pack CGI-compatible hex codes into actual non-word ASCII character
   51: 
   52: =item *
   53: 
   54: add_get_param() :
   55:  Inputs:  url (with or without exit GET from parameters), hash ref of
   56:               form name => value pairs
   57: 
   58:  Return: url with properly added the form name elements and values to the 
   59:          the url doing proper escaping of the values and joining with ? or &
   60:          as needed
   61: 
   62: =back
   63: 
   64: =cut
   65: 
   66: package LONCAPA;
   67: 
   68: use strict;
   69: use lib '/home/httpd/lib/perl/';
   70: use LONCAPA::Configuration;
   71: use Fcntl qw(:flock);
   72: use GDBM_File;
   73: use POSIX;
   74: 
   75: my $loncapa_max_wait_time = 13;
   76: 
   77: 
   78: use vars qw($match_domain   $match_not_domain
   79: 	    $match_username $match_not_username
   80: 	    $match_courseid $match_not_courseid
   81: 	    $match_name
   82:             $match_lonid
   83: 	    $match_handle   $match_not_handle);
   84: 
   85: require Exporter;
   86: our @ISA = qw (Exporter);
   87: our @EXPORT = qw(&add_get_param    &escape            &unescape       
   88: 		 &tie_domain_hash  &untie_domain_hash &tie_user_hash
   89: 		 &untie_user_hash  &propath);
   90: our @EXPORT_OK = qw($match_domain   $match_not_domain
   91: 		    $match_username $match_not_username
   92: 		    $match_courseid $match_not_courseid
   93: 		    $match_name
   94: 		    $match_lonid
   95: 		    $match_handle   $match_not_handle);
   96: our %EXPORT_TAGS = ( 'match' =>[qw($match_domain   $match_not_domain
   97: 				   $match_username $match_not_username
   98: 				   $match_courseid $match_not_courseid
   99: 				   $match_name
  100: 				   $match_lonid
  101: 				   $match_handle   $match_not_handle)],);
  102: my %perlvar;
  103: 
  104: 
  105: 
  106: # Inputs are a url, and a hash ref of
  107: # form name => value pairs
  108: # takes care of properly adding the form name elements and values to the 
  109: # the url doing proper escaping of the values and joining with ? or & as 
  110: # needed
  111: 
  112: sub add_get_param {
  113:     my ($url,$form_data) = @_;
  114:     my $needs_question_mark = ($url !~ /\?/);
  115: 
  116:     while (my ($name,$value) = each(%$form_data)) {
  117: 	if ($needs_question_mark) {
  118: 	    $url.='?';
  119: 	    $needs_question_mark = 0;
  120: 	} else { 
  121: 	    $url.='&';
  122: 	}
  123: 	$url.=$name.'='.&escape($form_data->{$name});
  124:     }
  125:     return $url;
  126: }
  127: 
  128: # -------------------------------------------------------- Escape Special Chars
  129: 
  130: sub escape {
  131:     my $str=shift;
  132:     $str =~ s/(\W)/"%".unpack('H2',$1)/eg;
  133:     return $str;
  134: }
  135: 
  136: # ----------------------------------------------------- Un-Escape Special Chars
  137: 
  138: sub unescape {
  139:     my $str=shift;
  140:     $str =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C",hex($1))/eg;
  141:     return $str;
  142: }
  143: 
  144: $match_domain     = $LONCAPA::domain_re     = qr{[\w\-.]+};
  145: $match_not_domain = $LONCAPA::not_domain_re = qr{[^\w\-.]+};
  146: sub clean_domain {
  147:     my ($domain) = @_;
  148:     $domain =~ s/$match_not_domain//g;
  149:     return $domain;
  150: }
  151: 
  152: $match_username     = $LONCAPA::username_re     = qr{\w[\w\-.@]+};
  153: $match_not_username = $LONCAPA::not_username_re = qr{[^\w\-.@]+};
  154: sub clean_username {
  155:     my ($username) = @_;
  156:     $username =~ s/^\W+//;
  157:     $username =~ s/$match_not_username//g;
  158:     return $username;
  159: }
  160: 
  161: 
  162: $match_courseid     = $LONCAPA::courseid_re     = qr{\d[\w\-.]+};
  163: $match_not_courseid = $LONCAPA::not_courseid_re = qr{[^\w\-.]+};
  164: sub clean_courseid {
  165:     my ($courseid) = @_;
  166:     $courseid =~ s/^\D+//;
  167:     $courseid =~ s/$match_not_courseid//g;
  168:     return $courseid;
  169: }
  170: 
  171: $match_name         = $LONCAPA::name_re = qr{$match_username|$match_courseid};
  172: sub clean_name {
  173:     my ($name) = @_;
  174:     $name =~ s/$match_not_username//g;
  175:     return $name;
  176: }
  177: 
  178: $match_lonid     = $LONCAPA::lonid_re     = qr{[\w\-.]+};
  179: 
  180: sub split_courseid {
  181:     my ($courseid) = @_;
  182:     my  ($domain,$coursenum) = 
  183: 	($courseid=~m{^/($match_domain)/($match_courseid)});
  184:     return ($domain,$coursenum);
  185: }
  186: 
  187: $match_handle     = $LONCAPA::handle_re     = qr{[\w\-.@]+};
  188: $match_not_handle = $LONCAPA::not_handle_re = qr{[^\w\-.@]+};
  189: sub clean_handle {
  190:     my ($handle) = @_;
  191:     $handle =~ s/$match_not_handle//g;
  192:     return $handle;
  193: }
  194: 
  195: # -------------------------------------------- Return path to profile directory
  196: 
  197: sub propath {
  198:     my ($udom,$uname)=@_;
  199:     $udom = &clean_domain($udom);
  200:     $uname= &clean_name($uname);
  201:     my $subdir=$uname.'__';
  202:     $subdir =~ s/(.)(.)(.).*/$1\/$2\/$3/;
  203:     my $proname="$perlvar{'lonUsersDir'}/$udom/$subdir/$uname";
  204:     return $proname;
  205: } 
  206: 
  207: 
  208: #---------------------------------------------------------------
  209: 
  210: =pod
  211: 
  212: =item tie_domain_hash()
  213: 
  214: Manipulation of hash based databases (factoring out common code
  215: for later use as we refactor.
  216: 
  217:  Ties a domain level resource file to a hash.
  218:  If requested a history entry is created in the associated hist file.
  219: 
  220:  Parameters:
  221:     domain    - Name of the domain in which the resource file lives.
  222:     namespace - Name of the hash within that domain.
  223:     how       - How to tie the hash (e.g. GDBM_WRCREAT()).
  224:     loghead   - Optional parameter, if present a log entry is created
  225:                 in the associated history file and this is the first part
  226:                  of that entry.
  227:     logtail   - Goes along with loghead,  The actual logentry is of the
  228:                 form $loghead:<timestamp>:logtail.
  229: Returns:
  230:    Reference to a hash bound to the db file or alternatively undef
  231:    if the tie failed.
  232: 
  233: =cut
  234: 
  235: sub tie_domain_hash {
  236:     my ($domain,$namespace,$how,$loghead,$logtail) = @_;
  237:     
  238:     # Filter out any whitespace in the domain name:
  239:     
  240:     $domain = &clean_domain($domain);
  241:     
  242:     # We have enough to go on to tie the hash:
  243:     
  244:     my $user_top_dir   = $perlvar{'lonUsersDir'};
  245:     my $domain_dir     = $user_top_dir."/$domain";
  246:     my $resource_file  = $domain_dir."/$namespace";
  247:     return &_locking_hash_tie($resource_file,$namespace,$how,$loghead,$logtail);
  248: }
  249: 
  250: sub untie_domain_hash {
  251:     return &_locking_hash_untie(@_);
  252: }
  253: 
  254: =pod
  255: 
  256: =item tie_user_hash()
  257: 
  258:   Ties a user's resource file to a hash.  
  259:   If necessary, an appropriate history
  260:   log file entry is made as well.
  261:   This sub factors out common code from the subs that manipulate
  262:   the various gdbm files that keep keyword value pairs.
  263: Parameters:
  264:   domain       - Name of the domain the user is in.
  265:   user         - Name of the 'current user'.
  266:   namespace    - Namespace representing the file to tie.
  267:   how          - What the tie is done to (e.g. GDBM_WRCREAT().
  268:   loghead      - Optional first part of log entry if there may be a
  269:                  history file.
  270:   what         - Optional tail of log entry if there may be a history
  271:                  file.
  272: Returns:
  273:   hash to which the database is tied.  It's up to the caller to untie.
  274:   undef if the has could not be tied.
  275: 
  276: =cut
  277: 
  278: sub tie_user_hash {
  279:     my ($domain,$user,$namespace,$how,$loghead,$what) = @_;
  280: 
  281:     $namespace=~s{/}{_}g;	# / -> _
  282:     $namespace     = &clean_username($namespace);
  283:     my $proname    = &propath($domain, $user);
  284:     my $file_prefix="$proname/$namespace";
  285:     return &_locking_hash_tie($file_prefix,$namespace,$how,$loghead,$what);
  286: }
  287: 
  288: sub untie_user_hash {
  289:     return &_locking_hash_untie(@_);
  290: }
  291: 
  292: =pod
  293: 
  294: =item locking_hash_tie()
  295: 
  296: routines if you just have a filename
  297: return tied hashref or undef
  298: 
  299: =cut
  300: 
  301: sub locking_hash_tie {
  302:     my ($filename,$how)=@_;
  303:     my ($file_prefix,$namespace)=&db_filename_parts($filename);
  304:     if ($namespace eq '') { return undef; }
  305:     return &_locking_hash_tie($file_prefix,$namespace,$how);
  306: }
  307: 
  308: sub locking_hash_untie {
  309:     return &_locking_hash_untie(@_);
  310: }
  311: 
  312: sub db_filename_parts {
  313:     my ($filename)=@_;
  314:     my ($file_path,$namespace)=($filename=~/^(.*)\/([^\/]+)\.db$/);
  315:     if ($namespace eq '') { return undef; }
  316:     return ($file_path.'/'.$namespace,$namespace);
  317: }
  318: 
  319: # internal routines that handle the actual tieing and untieing process
  320: 
  321: sub _do_hash_tie {
  322:     my ($file_prefix,$namespace,$how,$loghead,$what) = @_;
  323:     my %hash;
  324:     if(tie(%hash, 'GDBM_File', "$file_prefix.db", $how, 0640)) {
  325: 	# If this is a namespace for which a history is kept,
  326: 	# make the history log entry:    
  327: 	if (($namespace !~/^nohist\_/) && (defined($loghead))) {
  328: 	    my $hfh = IO::File->new(">>$file_prefix.hist"); 
  329: 	    if($hfh) {
  330: 		my $now = time();
  331: 		print $hfh ("$loghead:$now:$what\n");
  332: 	    }
  333: 	    $hfh->close;
  334: 	}
  335: 	return \%hash;
  336:     } else {
  337: 	return undef;
  338:     }
  339: }
  340: 
  341: sub _do_hash_untie {
  342:     my ($hashref) = @_;
  343:     my $result = untie(%$hashref);
  344:     return $result;
  345: }
  346: 
  347: {
  348:     my $sym;
  349:     my @pushed_syms;
  350: 
  351:     sub clean_sym {
  352: 	undef($sym);
  353:     }
  354:     sub push_locking_hash_tie {
  355: 	if (!defined($sym)) {
  356: 	    die("Invalid used of push_locking_hash_tie, should only be called after a lock has occurred and before and unlock.");
  357: 	}
  358: 	push(@pushed_syms,$sym);
  359: 	undef($sym);
  360:     }
  361: 
  362:     sub pop_locking_hash_tie {
  363: 	if (defined($sym)) {
  364: 	    die("Invalid nested used of pop_locking_hash_tie, should only be called after a unlock has occurred.");
  365: 	}
  366: 	$sym = pop(@pushed_syms);
  367:     }
  368: 
  369:     sub _locking_hash_tie {
  370: 	my ($file_prefix,$namespace,$how,$loghead,$what) = @_;
  371: 	if (defined($sym)) {
  372: 	    die('Nested locking attempted without proper use of push_locking_hash_tie, this is unsupported');
  373: 	}
  374: 
  375:         my $lock_type=LOCK_SH;
  376: # Are we reading or writing?
  377:         if ($how eq &GDBM_READER()) {
  378: # We are reading
  379:            if (!open($sym,"$file_prefix.db.lock")) {
  380: # We don't have a lock file. This could mean
  381: # - that there is no such db-file
  382: # - that it does not have a lock file yet
  383:                if ((! -e "$file_prefix.db") && (! -e "$file_prefix.db.gz")) {
  384: # No such file. Forget it.                
  385:                    $! = 2;
  386: 		   &clean_sym();
  387:                    return undef;
  388:                }
  389: # Apparently just no lock file yet. Make one
  390:                open($sym,">>$file_prefix.db.lock");
  391:            }
  392: # Do a shared lock
  393:            if (!&flock_sym(LOCK_SH)) { 
  394: 	       &clean_sym();
  395: 	       return undef; 
  396: 	   } 
  397: # If this is compressed, we will actually need an exclusive lock
  398: 	   if (-e "$file_prefix.db.gz") {
  399: 	       if (!&flock_sym(LOCK_EX)) {
  400: 		   &clean_sym();
  401: 		   return undef;
  402: 	       }
  403: 	   }
  404:         } elsif ($how eq &GDBM_WRCREAT()) {
  405: # We are writing
  406:            open($sym,">>$file_prefix.db.lock");
  407: # Writing needs exclusive lock
  408:            if (!&flock_sym(LOCK_EX)) {
  409: 	       &clean_sym();
  410: 	       return undef;
  411: 	   }
  412:         } else {
  413:            die("Unknown method $how for $file_prefix");
  414:         }
  415: # The file is ours!
  416: # If it is archived, un-archive it now
  417:        if (-e "$file_prefix.db.gz") {
  418:            system("gunzip $file_prefix.db.gz");
  419: 	   if (-e "$file_prefix.hist.gz") {
  420: 	       system("gunzip $file_prefix.hist.gz");
  421: 	   }
  422:        }
  423: # Change access mode to non-blocking
  424:        $how=$how|&GDBM_NOLOCK();
  425: # Go ahead and tie the hash
  426:       	my $result = 
  427: 	    &_do_hash_tie($file_prefix,$namespace,$how,$loghead,$what);
  428: 	if (!$result) {
  429: 	    &clean_sym();
  430: 	}
  431: 	return $result;
  432:     }
  433: 
  434:     sub flock_sym {
  435:         my ($lock_type)=@_;
  436: 	my $failed=0;
  437: 	eval {
  438: 	    local $SIG{__DIE__}='DEFAULT';
  439: 	    local $SIG{ALRM}=sub {
  440: 		$failed=1;
  441: 		die("failed lock");
  442: 	    };
  443: 	    alarm($loncapa_max_wait_time);
  444: 	    flock($sym,$lock_type);
  445: 	    alarm(0);
  446: 	};
  447: 	if ($failed) {
  448: 	    $! = 100; # throwing error # 100
  449: 	    return undef;
  450: 	} else {
  451: 	    return 1;
  452: 	}
  453:     }
  454: 
  455:     sub _locking_hash_untie {
  456: 	my ($hashref) = @_;
  457: 	my $result = untie(%$hashref);
  458: 	flock($sym,LOCK_UN);
  459: 	close($sym);
  460: 	&clean_sym();
  461: 	return $result;
  462:     }
  463: }
  464: 
  465: BEGIN {
  466:     %perlvar=%{&LONCAPA::Configuration::read_conf('loncapa.conf')};
  467: }
  468: 
  469: 1;
  470: 
  471: __END__
  472: 
  473: 

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