/) {next;} if($_ =~ /^<LINK REL/) {next;} if($_ =~ /^<\/HEAD>/) {next;} if($_ =~ /^<BODY/) {next;} if($_ =~ /^<\/BODY>/) {next;} if($_ =~ /^<\/HTML>/) {next;} if($_ =~ /^<BR> <HR>/) {next;} print TEXNEW ; } close(TEXNEW); ## Now do global multiline changes on whole file open(TEXNEW, "${TMPPROBDIR}${psvn}output.html") or die "Can't open ${TMPPROBDIR}${psvn}output.html"; @l2hOutputArray = <TEXNEW>; close(TEXNEW); my $l2hOutputString = join('',@l2hOutputArray); ## make gif images created by latex2html locatable by server ## NOTE: $htmlURL is defined in webworkCourse.ph . Often this will ## will be a link appearing in a public_html_docs directory. ## The $htmlURL, any links, and the next line must be coordinated. $l2hOutputString =~ s|${psvn}img|${PROBURL}${psvn}img|g; ## remove multiline comments $l2hOutputString =~ s|\n||sg; open(TEXNEW, ">${TMPPROBDIR}${psvn}output.html") or die "Can't open ${TMPPROBDIR}${psvn}output.html"; print TEXNEW $l2hOutputString; close(TEXNEW); ## remove unneeded files unless ($debug) {unlink("${TMPPROBDIR}${psvn}output.html.org");} unless ($debug) {unlink(<${TMPPROBDIR}*images.*>);} unless ($debug) {unlink(<${TMPPROBDIR}.*.db>);} unless ($debug) {unlink(<${TMPPROBDIR}*.db>);} unless ($debug) {unlink(<${TMPPROBDIR}IMG_PARAMS.*>);} unless ($debug) {unlink(<${TMPPROBDIR}*.pl>);} unless ($debug) {unlink(<${TMPPROBDIR}*.css>);} unless ($debug) {unlink("${TMPPROBDIR}index.html");} unless ($debug) {unlink("${TMPPROBDIR}${psvn}output.tex");} unless ($debug) {unlink("${TMPPROBDIR}${psvn}l2h.log");} unless ($debug) { my @allfiles = (); opendir( DIRHANDLE, "$TMPPROBDIR") || warn qq/Can't read directory $TMPPROBDIR $!/; @allfiles = map "$TMPPROBDIR$_", grep( /^l2h/, readdir DIRHANDLE); closedir(DIRHANDLE); my $l2hTempDir = $allfiles[0]; if (defined $l2hTempDir) { unlink(<$l2hTempDir/*>); rmdir ($l2hTempDir); } } ## change permission and group on remaining files chmod($Global::l2h_data_permission, glob("${TMPPROBDIR}*")); chown(-1,$Global::numericalGroupID,glob("${TMPPROBDIR}*")); ## Now that all the processing has been done, rename the $TMPPROBDIR TO $PROBDIR rename("$TMPPROBDIR","$PROBDIR") or warn "Can't rename the temporary problem directory:\n $TMPPROBDIR to $PROBDIR\n at displayMacros.pl , line: " . __LINE__ ; } ######################################################################################################### ##Subroutine that makes answers sticky in l2h mode # # # # INPUT: $rh_submittedAnswers Reference to a hash containing the answers submitted # # $ra_printLines Reference to an array containing the (HTML) text to be output # # $rh_flags Reference to a hash containing flags; specifically a # # reference to an array containing the answer field labels # # # # OUTPUT: @printLines An array containing the (modified) text to be output # # # # OVERVIEW: l2h_sticky_answers is given HTML text, a list of submitted answers, and a list of # # answer field labels. Its job is to retain the user's answers between submissions # # when in typeset mode (this is handled elsewhere in the text modes). Basically, its # # job is to act as a "filter" for the HTML text, replacing the answer fields that have # # been reset with fields containing the previously entered answers, returning the # # modified text. A brief high-level overview of the algorithm follows: # # # # ALGORITHM: The references are first dereferenced. The incoming text is first joined into # # one string. It is then split up again, but not by line. Rather, the text is split # # such that each array entry is either text which can be ignored, or a single # # <INPUT...> tag. Each entry is then processed. If it is an <INPUT> tag, then it # # must be checked for the presence of each answer field label for which a value was # # submitted (there are many <INPUT> fields which are not answer fields, so we can't # # assume that consecutive <INPUT> fields correspond to consecutive answer labels). # # If a label is found, the blank value space is replaced with the appropriate # # submitted answer (note that we can assume that there is a one-to-one correspondence # # between answer labels and submitted answers; this is guaranteed by the specs). Radio # # buttons and checkboxes are handled specially; see below. The modified text is then # # added to the output string, which is split on a placeholder such that the output # # array has the same number of entries as the input array (this is not required, but # # might avoid some subtle bug in the future). # # # # NOTE: The specifications seem to require that the input text array consist of one # # field for each line of text. However, it appears that the input is actually one # # field, with newline characters separating lines. This function should accept # # either form of input, although the "correct" form of one field per line has not # # been tested. It is possible that, if input is received in this form AND the # # newline characters have been truncated, the output could be garbled. # # # # --David Etlinger 6/7/2000 # # # # ADDED: Added a few lines of code to properly handle radio buttons. Checkboxes still need # # to be implemented. # # --David Etlinger 6/14/2000 # # # # ADDED: Added code to handle checkboxes. This is complicated because the submitted checkboxes # # are originally stored as a single string with "\0" as a delimiter. If the input type # # is determined to be checkboxes, the string is first split into an array. A hash key # # in a special checkbox array is then made to point to the array. This is done because # # there might be more than one checkbox set in a single question. Each time an input line # # of type checkbox appears, the next value in this array is popped into a temp variable. # # If it is determined that the line being processed corresponds to this value, the line # # is processed (made "sticky"); otherwise, the value is pushed back on the array. The # # fact that the number of checked cehckboxes is known but the total number of checkboxes # # is not means that a given line of input type checkbox might or might not correspond # # to the next value in the checkbox array. (I hope this explanation is clear enough!) # # --David Etlinger 6/28/2000 # ######################################################################################################### sub l2h_sticky_answers { my ( $rh_submittedAnswers, $ra_printLines, $rh_flags ) = @_; #warn ("rh_submittedAnswers = \@rh_submittedAnswers"); #warn ("ra_printLines = \@{ra_printLines}"); #warn ("rh_flags = \@{rh_flags}"); my %submittedAnswers = %{$rh_submittedAnswers}; my @printLines = @{$ra_printLines}; my @answerLabels = @{$rh_flags -> {ANSWER_ENTRY_ORDER}}; my $line; # holds the text of each line my $label; # holds each answer label my $counter = 0; # holds the index of the current answer my $output; # holds the text the subroutine returns my $answer_value; my %checkboxAns; # holder for the checkbox multi-part answers my $nextCheckboxAns; # temp holder for the next checkbox answer to be processed my $placeholder = "\x253"; # unused hex character to join text lines with #first, convert the array of text lines to one string... my $text = join( "$placeholder", @printLines ); #then, split it such that a line consists of either text #or a single <INPUT> tag (case insensitive; note also that #whitespace within the <INPUT> tag is accounted for). # NOTE -- the regular expression searches for "<", then any # amount of whitespace, then "INPUT", then any number of # characters that aren't ">", then ">". I think that instead of # searching for characters that aren't ">", I could have instead # searched to match a minimal number of characters (using ?), and # then ">". I don't know regular expressions well enough to tell # if this might lead to some subtle difference. my @textLines = split( m|(<\s*INPUT[^>]*>)|is, $text ); #my @textLines = split( m|(<\s*INPUT.*?>)|is, $text ); foreach $line ( @textLines ) { if( $line =~ m|<\s*INPUT|i ) { foreach $label ( @answerLabels ) { next unless exists( $submittedAnswers{$label} ); # skip if no answer was submitted. if( $line =~ m|NAME\s*=\s*"$label"|i ) { if( $line =~ m|TYPE\s*=\s*RADIO|i ) { #handle radio buttons $line =~ s|VALUE\s*=\s*"$submittedAnswers{$label}"|VALUE = "$submittedAnswers{$label}" CHECKED|i; } elsif( $line =~ m|TYPE\s*=\s*CHECKBOX|i ) { #make the hash key point to an anonymous array $checkboxAns{$label} = [ split( "\0", $submittedAnswers{$label} ) ] if not exists( $checkboxAns{$label} ); if( defined $checkboxAns{$label}[0] ) { $nextCheckboxAns = shift @{$checkboxAns{$label}}; if( $line !~ s|VALUE\s*=\s*"$nextCheckboxAns"|VALUE = "$nextCheckboxAns" CHECKED|i ) { unshift( @{$checkboxAns{$label}}, $nextCheckboxAns ); #put the unused answer back on the list } } } else { # we'll assume this is something else, like one or more fields. # if it's several fields, we need to take only one answer at a time # \0 are used to delimeter between entries. if ($submittedAnswers{$label} =~ /\0/ ) { my @answers = split("\0", $submittedAnswers{$label}); $answer_value = shift(@answers); # use up the first answer $submittedAnswers{$label}=join "\0", @answers; # store the rest $answer_value= '' unless defined($answer_value); } else { $answer_value = $submittedAnswers{$label}; } $line =~ s|VALUE\s*=\s*""|VALUE = "$answer_value"|i; } } } } #end if test for "<INPUT" $output .= $line; } #end foreach @printLines = split( m|$placeholder|, $output ); return @printLines; } #end l2h_sticky_answers() ## This is the old system (but newer than the one below). ## It has been replaced for two reasons: ## 1) It is complicated and difficult to understand or modify ## 2) It does not work for several situations that rarely come up, ## but must be handled properly. Specifically, it doesn't handle ## text with more than one <INPUT> tag on a given line very well. ## there are probably other problems, but that is the biggest. ## --DME 6/7/2000 # # the following doubly nested loop iterates over each line, # # and for each line searches for each answer label. Technically, # # it might have been faster to join each entry in @printlines # # into one string, search on that, and split it back up, but I # # felt that the slight theoretical speed gain was not worth the # # added complexity. # warn "answerLabels = @answerLabels"; #DEBUG # foreach $line ( @printLines ) { # warn "Line is $line"; #DEBUG # foreach $label ( @answerLabels ) { # if( $line =~ m|<INPUT TYPE=TEXT.*NAME="$label| ) { # while ($line =~ /VALUE = ""/) { # # Put trailing space in displayed answer so that while loop will # # always end. We are using the form of the s/// operator which # # evaluates its right hand side # $line =~ s|NAME="$label" VALUE = ""| # $counter++; # $submittedAnswers[$counter]=" " unless defined ($submittedAnswers[$counter]) # && not $submittedAnswers[$counter] =~ /^\s*$/; # qq{ NAME="$label" VALUE = "$submittedAnswers[$counter]" } |e; # # This insures that in VALUE = "$submittedAnswers[$counter]" # # the quantity $submittedAnswers[$counter] # # is never empty. This is required in order to terminate the loop. # } #end while # push( @output, $line ); # } #end if # else { # push( @output, $line ); # } # } #end foreach over @answerLabels # } #end foreach over @printLines # # @printLines = @output; # } #end outer if # # return @printLines; # } #end l2h_sticky_answers() ##subroutine that makes answers sticky in l2h mode # this is an old version of this routine, which assumes (incorrectly) # that answer labels begin with "AnSwEr". I've left it here just in case... # DME 6/6/2000 #sub l2h_sticky_answers { # my ($refSubmittedAnswers, $refprintlines)=@_; # my @printlines=@$refprintlines; # if ((@{$refSubmittedAnswers}!=0)) { # my $line; # my @output=(); # foreach $line (@printlines) { # if ($line =~ m|<INPUT TYPE=TEXT.*NAME="AnSwEr|) { # #print "<P>line doesn't exists<P>\n" unless defined($line); # while ($line =~ /VALUE = ""/) { # ## Put trailing space in displayed answer so that while loop will # ## always end. We are using the form of the s/// operator which evaluates its right hand side # $line =~ s|NAME="AnSwEr(\d*)" VALUE = ""| # my $tttemp = $1; # ${$refSubmittedAnswers}[$tttemp-1]=" " unless defined (${$refSubmittedAnswers}[$tttemp-1]) # && not ${$refSubmittedAnswers}[$tttemp-1] =~ /^\s*$/; # # qq{ NAME="AnSwEr$tttemp" VALUE = "${$refSubmittedAnswers}[$tttemp-1]" } |e; # # This insures that in VALUE = "${$refSubmittedAnswers}[$tttemp-1]" the quantity ${$refSubmittedAnswers}[$tttemp-1] # # is never empty. This is required in order to terminate the loop. # } # push(@output, $line); # } # else { # push(@output, $line); # } # } # # @printlines = @output; # } # # @printlines; #} ##subroutine that updates current keys in the l2h mode # sub l2h_update_keys { # my ($sessionKey, $refprintlines)= @_; # my @printlines=@$refprintlines; # my $line; # my @output=(); # #my $sessionKey = $main::sessionKey; # warn "hi lines = ",join("",@printlines); # foreach $line (@printlines) { # if ($line =~ m|^\s*<A(.*?)\&key=[^&]*&user|) { #<A.*&key=.*?&user # #grab the session key from the CGI input or make it blank # $line =~ s|^\s*<A(.*?)&key=[^&]*&user|<A$1&key=$sessionKey&user|; # warn "line = $line<BR>"; # push(@output, $line); # }else{ # push(@output, $line); # } # # } # @printlines; # # } sub makeL2H { my ($TMPPROBDIR,$psvn) =@_; $ENV{PATH} .= "$Global::extendedPath"; if($Global::externalLaTeX2HTMLVersion eq "98.1p1") { system("$Global::externalLaTeX2HTMLPath -no_math -init_file $Global::externalLaTeX2HTMLInit -dir $TMPPROBDIR -prefix $psvn $TMPPROBDIR${psvn}output.tex > $TMPPROBDIR${psvn}l2h.log 2>&1"); } elsif($Global::externalLaTeX2HTMLVersion eq "96.1") { system("$Global::externalLaTeX2HTMLPath -init_file $Global::externalLaTeX2HTMLInit -dir $TMPPROBDIR -prefix $psvn $TMPPROBDIR${psvn}output.tex > $TMPPROBDIR${psvn}l2h.log"); } else { die "Unknown LaTeX2HTML version: \$Global::externalLaTeX2HTMLVersion = $Global::externalLaTeX2HTMLVersion"; } } 1;

=head1 displayMacros.pl This file is used with WeBWorK 1.9 and is not used for WeBWorK 2.x. =cut use strict; ## $ENV{'PATH'} .= ':/usr/math/bin'; my $debug = 0; $debug = 1 if $Global::imageDebugMode; ## if $debug =1, log, etc. files created by ## latex2html are not deleted ############################################################## # File: DisplayMacros.pl # This contains the subroutines for creating problem files ############################################################## ################################################################ # Copyright @1995-1998 by Michael E. Gage, Arnold K. Pizer and # WeBWorK at the University of Rochester. All rights reserved. ################################################################ ## To add or delete displayModes edit this file sub displaySelectModeLine_string # called from probSet.pl # displays the option line for selecting display modes { my ($displayMode) =@_ ; $displayMode = $Global::htmlModeDefault unless(defined($displayMode)); # If the system is set up with only one display mode, there is # no need to display a choice - use the default if(scalar(@{$Global::available_mode_list})<2) { return(''); } my $out = "Display Mode:
"; # A list of the available modes. my $mode_list = $Global::available_mode_list; ## ref to a list of available modes ## The format is [internal symbol, external name] # A list of the available modes. # Format is [internal symbol, external name, ""], where the third # argument is changed to checked below for the current displayMode # my $mode_list = [ # ['HTML', 'text', ""], # ['HTML_tth', 'formatted-text',""], # ['HTML_dpng' ,'dvipng',""], # ['Latex2HTML', 'typeset',""] # ]; # Make the format [internal symbol, external name, ''] # The third argument is changed to checked below for the current displayMode my $j; for $j (0..(scalar(@{$mode_list})-1)) { push @{$mode_list->[$j]},''; } if (! defined($displayMode) ) {$displayMode = $Global::htmlModeDefault;} my $found = 0; # Search through all modes to match for displayMode # If we don't find one, found=0 will trigger warn message below for $j (0..(scalar(@{$mode_list})-1)) { if($mode_list->[$j]->[0] eq $displayMode) { $mode_list->[$j]->[2] = "CHECKED"; $found=1; last; } } for $j (@{$mode_list}) { $out .= qq![2]>$j->[1]
\n!; } if(! $found) { my $wstr = " Error: displayMacros.pl: sub displaySelectModeLine. Unrecognized mode |$displayMode| . The acceptable modes are: "; for $j (@{$mode_list}) { $wstr .= " $j->[0] "; } warn $wstr; } $out; } sub displaySelectModeLine { print displaySelectModeLine_string(@_); } ################################################################################################################## # Does the initial processing of the problem. # Returns an array containing the rendered problem. # ################################################################################################################## sub createDisplayedProblem { my ($setNumber,$probNum,$psvn,$printlinesref,$rh_flags)= @_; my @printlines; my $coursel2hDirectory = getCoursel2hDirectory(); unless(-e $coursel2hDirectory ) { &createDirectory($coursel2hDirectory, $Global::l2h_set_directory_permission, $Global::numericalGroupID); } unless(-e "${coursel2hDirectory}set$setNumber") { &createDirectory("${coursel2hDirectory}set$setNumber",$Global::l2h_set_directory_permission, $Global::numericalGroupID); } my $PROBDIR = convertPath("${coursel2hDirectory}set$setNumber/$probNum-$psvn/"); my $TMPPROBDIR = convertPath("${coursel2hDirectory}$probNum-$psvn/"); if (! -e $PROBDIR) { # no gifs of equations have been created &l2hcreate($setNumber,$probNum,$psvn,$printlinesref); } else { # determine if the gifs are older than the modifications of the source file #&attachProbSetRecord($psvn); my $fileName = &getProblemFileName($probNum,$psvn); $fileName = "${Global::templateDirectory}$fileName"; #print "\n\n The filename is $fileName \n\n"; my @probDirStat = stat $PROBDIR; my @sourceFileStat = stat $fileName; #print "\n\n The source file age is $sourceFileStat[9] \n\n"; #print "\n\n The prob dir age is $probDirStat[9] \n\n"; if (($sourceFileStat[9] > $probDirStat[9] ) or $rh_flags->{'refreshCachedImages'}) { ## source file is newer or solutions should be shown recreate the l2h cache rmDirectoryAndFiles($PROBDIR); &l2hcreate($setNumber,$probNum,$psvn,$printlinesref); } } #the problem has been rendered by Latex2HTML into this file: # open(TEXXX, "${PROBDIR}${psvn}output.html") || die "Can't open ${PROBDIR}${psvn}output.html"; open(TEXXX, "${PROBDIR}${psvn}output.html") or warn "ERROR: $0". "Can't open the HTML file: \n ${PROBDIR}${psvn}output.html\n(allegedly)". "translated by latex2HTML\n at displayMacros.pl, line" . __LINE__ ; @printlines = ; push(@printlines, "The file ${PROBDIR}${psvn}output.html was empty") unless @printlines; #print "PRINTLINES",@printlines; close(TEXXX); @printlines; } ########################################################################################### # Formats and displays the responses to submitted answers to the problem. Returns a string. # ########################################################################################### sub display_answers { # this will be put in displayMacros.pl soon. #my ($displayCorrectAnswersQ,$showPartialCorrectAnswers,$rh_answer_results,$rh_problem_result) = @_; my ($rh_answer_results,$rh_problem_result,$rh_flags) = @_; my $displayCorrectAnswersQ = $rh_flags ->{displayCorrectAnswersQ}; my $showPartialCorrectAnswers = $rh_flags -> {showPartialCorrectAnswers}; my @answer_entry_order = @{$rh_flags -> {ANSWER_ENTRY_ORDER} }; my $ANSWER_PREFIX = $rh_flags -> {ANSWER_PREFIX}; my $allAnswersCorrectQ = 1; my $printedResponse=''; ###### Print appropriate response to submitted answers my ($i,$answerIsCorrectQ, $normalizedSubmittedAnswer,$normalizedCorrectAnswer,$ans_name,$errors); $i=0; # $printedResponse .= "\n\n"; # replace above line by next two lines as per Davide Cervone. AKP. $printedResponse .= "\n

\n"; $printedResponse .= "\n

\n"; foreach my $key ( @answer_entry_order ) { $i++; $answerIsCorrectQ = $rh_answer_results ->{$key} -> {score}; $normalizedSubmittedAnswer = $rh_answer_results ->{$key} -> {student_ans}; $normalizedSubmittedAnswer = '' if ($normalizedSubmittedAnswer =~ /^error:\s+empty/); $normalizedCorrectAnswer = $rh_answer_results ->{$key} -> {original_correct_ans}; ## Handle the case where the answer evaluator does not return original_correct_ans if ((!defined $normalizedCorrectAnswer) or (!$normalizedCorrectAnswer =~ /\S/)) { $normalizedCorrectAnswer = $rh_answer_results ->{$key} -> {correct_ans}; } $errors = $rh_answer_results ->{$key} -> {ans_message}; $errors = '' if ($errors eq 'empty'); #$ans_name = $rh_answer_results ->{$key} -> {ans_name}; #$ans_name =~ s/$ANSWER_PREFIX//; # this handles implicitly defined answer names. $ans_name = $i; # just number the answers in order $allAnswersCorrectQ = $allAnswersCorrectQ && $answerIsCorrectQ; $printedResponse .= "\n" if ($answerIsCorrectQ && $showPartialCorrectAnswers ); $printedResponse .= "Incorrect. " if (!($answerIsCorrectQ) && $showPartialCorrectAnswers); $errors =~ s/\n/
/g; ## convert newlines to
in error messages as per Davide Cervone # change 9/2/00 by MEG -- give width in pixels rather than %. # Some browsers break with % widht which is not the standard $printedResponse .= "\n" if ($errors =~ /\w/); $printedResponse .= "\n" if ($displayCorrectAnswersQ); } if ($i == 1) { $printedResponse .= "\n

Answer $ans_name entered:--> $normalizedSubmittedAnswer <-- "; $printedResponse .= "Correct.
	$errors
	Correct answer: $normalizedCorrectAnswer
The above answer is correct. " if ($allAnswersCorrectQ); $printedResponse .= "\n
The above answer is NOT correct. " if (!($allAnswersCorrectQ)); } else { $printedResponse .= "\n
All of the above answers are correct. " if ($allAnswersCorrectQ); $printedResponse .= "\n
At least one of the above answers is NOT correct. " if (!($allAnswersCorrectQ)); } my $percentCorr = int(100*$rh_problem_result->{score} +.5); $printedResponse .="\n
Your score on this attempt is ${percentCorr}\%. "; # $printedResponse .= "\n

\n"; # replace above line by next line as per Davide Cervone. AKP. $printedResponse .= "

\n\n"; # $printedResponse .="\n problem grader is ".$rh_problem_result->{type}." and the score is ".$rh_problem_result->{score}."
\n"; $printedResponse; } ########################################################################################### # Previews submitted answers to the problem. Returns a string. # ########################################################################################### sub preview_answers { my ($rh_answer_results,$rh_problem_result,$rh_flags) = @_; my @answer_entry_order = @{$rh_flags -> {ANSWER_ENTRY_ORDER} }; my $ANSWER_PREFIX = $rh_flags -> {ANSWER_PREFIX}; my $printedResponse =''; ###### Print appropriate response to submitted answers my ($i,$original_student_ans,$normalizedSubmittedAnswer,$errors,$ans_name,$preview_text_string,$preview_latex_string); my ($ans_evaluator_type, $value_word, $error_word, $show_value); $i=0; $printedResponse .= "\n\n"; foreach my $key ( @answer_entry_order ) { $i++; $ans_name = $rh_answer_results ->{$key} -> {ans_name}; #$ans_name =~ s/$ANSWER_PREFIX//; # this handles implicitly defined answer names. #commented out by DME 6/6/2000 $original_student_ans = $rh_answer_results ->{$key} -> {original_student_ans}; $normalizedSubmittedAnswer = $rh_answer_results ->{$key} -> {student_ans}; $errors = $rh_answer_results ->{$key} -> {ans_message}; $errors =~ s/\n/
/g; ## convert newlines to
in error messages as per Davide Cervone $preview_text_string =''; $preview_text_string = $rh_answer_results ->{$key} -> {preview_text_string} if defined $rh_answer_results ->{$key} -> {preview_text_string}; $preview_latex_string =''; $preview_latex_string = $rh_answer_results ->{$key} -> {preview_latex_string} if defined $rh_answer_results ->{$key} -> {preview_latex_string}; $ans_evaluator_type = $rh_answer_results ->{$key} -> {type}; $value_word = 'value:'; $show_value = 0; $show_value = 1 if ((($ans_evaluator_type =~ /number/) and ($normalizedSubmittedAnswer =~ /\w/)) or ($normalizedSubmittedAnswer =~ /^error/)); $show_value = 0 if ($normalizedSubmittedAnswer =~ /^error:\s+empty/); $value_word = '' if ($normalizedSubmittedAnswer =~ /^error/); $error_word = 'error:'; $error_word = '' if ($errors =~ /^error:/); $printedResponse .= "\n"; #$printedResponse .= "\n"; #commented out by DME 6/6/2000 $printedResponse .= "\n"; $printedResponse .= "\n" if ($preview_text_string =~ /\w/); $printedResponse .= "\n" if $show_value == 1; $printedResponse .= "\n" if (($errors =~ /\w/) and ($errors ne 'empty')) ; if ($preview_latex_string =~ /\w/) { $printedResponse .= "\n "; } $printedResponse .= "\n"; } $printedResponse .= "\n

Ans $i

	parsed: $preview_text_string
	${value_word} $normalizedSubmittedAnswer
	${error_word} $errors
	"; $printedResponse .= "\n "; $printedResponse .= "\n

\n"; $printedResponse; } sub lc_sort { # this sorts strings with letters and number groups, alternately lexigraphically and numerically # (lc stands for library of congress as in QA617.34R45) my($left,$right) = @_; # format "abcd345.57def34ABC"; # string assumed to begin with alpha # string is split into alternating alpha and numeric groups # numeric groups match [\d\.]+ # numeric groups assumed to contain at least one digit, ( a period alone will cause and error) # alpha groups can contain any characters except digits and the period # spaces in alpha groups will cause unexpected behavior # sort is not case sensitive # _ sorts after alpha characters # not case sensitive my @a = split( /([\d\.]+)/, $left); my @b = split( /([\d\.]+)/, $right); my $out = undef; my $mode = 0; # even is lexic and odd is numeric my($l,$r); while (@a) { $l = shift @a; $r = shift @b; $out = ($mode++ % 2 == 0) ? uc($l) cmp uc($r) : $l <=> $r; # lexic or numeric compare last unless $out==0; # stop unless $l and $r are different. } $out; } ##################################################################### # Creates an insert which appears on the probSet page. # ##################################################################### sub createDisplayedInsert { #my ($mode,$setNumber,$fileName,$psvn,$courseName,$printlinesref)= @_; my ($setNumber,$fileName,$psvn,$courseName,$printlinesref)= @_; my @printlines=@$printlinesref; my $PROBDIR; # if($mode eq "HTML" || $mode eq 'HTML_tth') { # @printlines = &createProblem2($mode,$fileName,$psvn,$courseName,$sourceref); # # } elsif ($mode eq 'Latex2HTML') { #latex2html processing my $coursel2hDirectory = getCoursel2hDirectory(); unless(-e $coursel2hDirectory ) { &createDirectory($coursel2hDirectory, $Global::l2h_set_directory_permission, $Global::numericalGroupID); } unless(-e "${coursel2hDirectory}set$setNumber") { &createDirectory("${coursel2hDirectory}set$setNumber",$Global::l2h_set_directory_permission, $Global::numericalGroupID); } my $shortFileName = $fileName; $shortFileName =~ s|^.*?([^\/]*)$|$1|; $shortFileName =~ s|\..*$||; $PROBDIR = convertPath("${coursel2hDirectory}set$setNumber/$shortFileName-$psvn/"); if (! -e $PROBDIR) { &l2hcreate($setNumber,$shortFileName,$psvn,$printlinesref); } else { #&attachProbSetRecord($psvn); my $fullFileName = "${Global::templateDirectory}$fileName"; #print "\n\n The full filename is $fullFileName \n\n"; my @probDirStat = stat $PROBDIR; my @sourceFileStat = stat $fullFileName; #print "\n\n The source file age is $sourceFileStat[9] \n\n"; #print "\n\n The prob dir age is $probDirStat[9] \n\n"; if ($sourceFileStat[9] > $probDirStat[9] ) { ## source file is newer rmDirectoryAndFiles($PROBDIR); &l2hcreate($setNumber,$shortFileName,$psvn,$printlinesref); } #else {&createProblem2($mode, $fileName, $psvn,$courseName,$sourceref);} ##initialize problem } open(TEXXX, "${PROBDIR}${psvn}output.html") or die "ERROR: $0 Can't open ${PROBDIR}${psvn}output.html"; @printlines = ; close(TEXXX); # } else { # # @printlines="createDisplayedProblem: Error: Mode is not HTML, HTML_tthHTML_tth or Latex2HTML."; # # # } @printlines; } ##do not need this subroutine anymore #sub l2hcreateProb { # my ($setNumber,$probNum,$psvn,$printlinesref)= @_; # #my ($setNumber,$probNum,$psvn,$courseName,$printlinesref)= @_; # #my $mode = 'Latex2HTML'; # # #my @printlines = &createProblem($mode, $probNum, $psvn, $courseName,$sourceref,$refSubmittedAnswers); # #my $printlinesref = \@printlines; # my $tmpDirectory = "tmp/l2h/set$setNumber/$probNum-$psvn/"; # l2hcreate($setNumber,$probNum,$psvn,$printlinesref) #} #do not use this subroutine anymore #sub l2hcreateInsert { # my ($setNumber,$shortFileName,$psvn,$printlinesref)= @_; # #my $mode = 'Latex2HTML'; # #my @printlines = &createProblem2($mode, $fileName, $psvn,$courseName,$sourceref); # #my $printlinesref = \@printlines; # #my $shortFileName = $fileName; # #$shortFileName =~ s|^.*?([^\/]*)$|$1|; # #my $tmpDirectory = "tmp/l2h/set$setNumber/$shortFileName-$psvn/"; # l2hcreate($setNumber,$shortFileName,$psvn,$printlinesref) #} sub l2hcreate { ## for latex2HTML 96.1 and 98.1 my ($setNumber,$probNum,$psvn,$printlinesref) = @_; # warn "l2hcreate is being executed displaymacros.pl line ".__LINE__; my $PROBDIR = convertPath(&getCoursel2hDirectory."set$setNumber/$probNum-$psvn/"); my $TMPPROBDIR = convertPath(&getCoursel2hDirectory."$probNum-$psvn/"); my $PROBURL = &getCoursel2hURL."set$setNumber/$probNum-$psvn/"; &createDirectory($TMPPROBDIR,$Global::l2h_prob_directory_permission,$Global::numericalGroupID) unless(-e "$TMPPROBDIR"); open(OUTTEXFILE, ">$TMPPROBDIR${psvn}output.tex") or die "Can't open temporary file $TMPPROBDIR${psvn}output.tex"; print OUTTEXFILE &texInput($Global::TEX_PROB_PREAMBLE); print OUTTEXFILE &texInput($Global::TEX_PROB_HEADER); print OUTTEXFILE @$printlinesref; print OUTTEXFILE &texInput($Global::TEX_PROB_FOOTER); close(OUTTEXFILE); ## Give this temporary file permission 666 in case the process dies before it it deleted 60 lines further down chmod(0666, "$TMPPROBDIR${psvn}output.tex"); ## system("/usr/math/bin/latex2html -init_file ${Global::mainDirectory}latex2html.init -dir $PROBDIR -prefix $psvn ${htmlDirectory}tmp/l2h/${psvn}output.tex > ${htmlDirectory}tmp/l2h/${psvn}l2h.log"); my $latex2HTML_result = &makeL2H($TMPPROBDIR, $psvn) ; warn( "LaTeX2HTML failed. Returned with status: $latex2HTML_result\n" ) if $latex2HTML_result ; ##Get rid of all unwanted stuff in html document created by latex2html unless(-e "${TMPPROBDIR}${psvn}output.html") { warn "Can't rename ${TMPPROBDIR}${psvn}output.html"; return (0); ### there was a failure in latex2html processing ### we just give a warning so that so that l2hPrecreateSet.pl can continue } rename("${TMPPROBDIR}${psvn}output.html","${TMPPROBDIR}${psvn}output.html.org") or warn "Can't rename ${TMPPROBDIR}${psvn}output.html at ". __LINE__; open(TEXORG, "${TMPPROBDIR}${psvn}output.html.org") or warn "Can't open ${TMPPROBDIR}${psvn}output.html.org"; my @l2hOutputArray; BLK: { # This is protection to make absolutely sure that the line separater is set properly. # It's still a mystery as to where this becomes defined to be something else. local($/); $/ = "\n"; @l2hOutputArray = ; } close(TEXORG); open(TEXNEW, ">${TMPPROBDIR}${psvn}output.html") or die "Can't open ${TMPPROBDIR}${psvn}output.html"; foreach (@l2hOutputArray) { if($_ =~ /^/) {next;} if($_ =~ /^/) {next;} if($_ =~ /^/) {next;} if($_ =~ /^<LINK REL/) {next;} if($_ =~ /^<\/HEAD>/) {next;} if($_ =~ /^<BODY/) {next;} if($_ =~ /^<\/BODY>/) {next;} if($_ =~ /^<\/HTML>/) {next;} if($_ =~ /^<BR> <HR>/) {next;} print TEXNEW ; } close(TEXNEW); ## Now do global multiline changes on whole file open(TEXNEW, "${TMPPROBDIR}${psvn}output.html") or die "Can't open ${TMPPROBDIR}${psvn}output.html"; @l2hOutputArray = <TEXNEW>; close(TEXNEW); my $l2hOutputString = join('',@l2hOutputArray); ## make gif images created by latex2html locatable by server ## NOTE: $htmlURL is defined in webworkCourse.ph . Often this will ## will be a link appearing in a public_html_docs directory. ## The $htmlURL, any links, and the next line must be coordinated. $l2hOutputString =~ s|${psvn}img|${PROBURL}${psvn}img|g; ## remove multiline comments $l2hOutputString =~ s|\n||sg; open(TEXNEW, ">${TMPPROBDIR}${psvn}output.html") or die "Can't open ${TMPPROBDIR}${psvn}output.html"; print TEXNEW $l2hOutputString; close(TEXNEW); ## remove unneeded files unless ($debug) {unlink("${TMPPROBDIR}${psvn}output.html.org");} unless ($debug) {unlink(<${TMPPROBDIR}*images.*>);} unless ($debug) {unlink(<${TMPPROBDIR}.*.db>);} unless ($debug) {unlink(<${TMPPROBDIR}*.db>);} unless ($debug) {unlink(<${TMPPROBDIR}IMG_PARAMS.*>);} unless ($debug) {unlink(<${TMPPROBDIR}*.pl>);} unless ($debug) {unlink(<${TMPPROBDIR}*.css>);} unless ($debug) {unlink("${TMPPROBDIR}index.html");} unless ($debug) {unlink("${TMPPROBDIR}${psvn}output.tex");} unless ($debug) {unlink("${TMPPROBDIR}${psvn}l2h.log");} unless ($debug) { my @allfiles = (); opendir( DIRHANDLE, "$TMPPROBDIR") || warn qq/Can't read directory $TMPPROBDIR $!/; @allfiles = map "$TMPPROBDIR$_", grep( /^l2h/, readdir DIRHANDLE); closedir(DIRHANDLE); my $l2hTempDir = $allfiles[0]; if (defined $l2hTempDir) { unlink(<$l2hTempDir/*>); rmdir ($l2hTempDir); } } ## change permission and group on remaining files chmod($Global::l2h_data_permission, glob("${TMPPROBDIR}*")); chown(-1,$Global::numericalGroupID,glob("${TMPPROBDIR}*")); ## Now that all the processing has been done, rename the $TMPPROBDIR TO $PROBDIR rename("$TMPPROBDIR","$PROBDIR") or warn "Can't rename the temporary problem directory:\n $TMPPROBDIR to $PROBDIR\n at displayMacros.pl , line: " . __LINE__ ; } ######################################################################################################### ##Subroutine that makes answers sticky in l2h mode # # # # INPUT: $rh_submittedAnswers Reference to a hash containing the answers submitted # # $ra_printLines Reference to an array containing the (HTML) text to be output # # $rh_flags Reference to a hash containing flags; specifically a # # reference to an array containing the answer field labels # # # # OUTPUT: @printLines An array containing the (modified) text to be output # # # # OVERVIEW: l2h_sticky_answers is given HTML text, a list of submitted answers, and a list of # # answer field labels. Its job is to retain the user's answers between submissions # # when in typeset mode (this is handled elsewhere in the text modes). Basically, its # # job is to act as a "filter" for the HTML text, replacing the answer fields that have # # been reset with fields containing the previously entered answers, returning the # # modified text. A brief high-level overview of the algorithm follows: # # # # ALGORITHM: The references are first dereferenced. The incoming text is first joined into # # one string. It is then split up again, but not by line. Rather, the text is split # # such that each array entry is either text which can be ignored, or a single # # <INPUT...> tag. Each entry is then processed. If it is an <INPUT> tag, then it # # must be checked for the presence of each answer field label for which a value was # # submitted (there are many <INPUT> fields which are not answer fields, so we can't # # assume that consecutive <INPUT> fields correspond to consecutive answer labels). # # If a label is found, the blank value space is replaced with the appropriate # # submitted answer (note that we can assume that there is a one-to-one correspondence # # between answer labels and submitted answers; this is guaranteed by the specs). Radio # # buttons and checkboxes are handled specially; see below. The modified text is then # # added to the output string, which is split on a placeholder such that the output # # array has the same number of entries as the input array (this is not required, but # # might avoid some subtle bug in the future). # # # # NOTE: The specifications seem to require that the input text array consist of one # # field for each line of text. However, it appears that the input is actually one # # field, with newline characters separating lines. This function should accept # # either form of input, although the "correct" form of one field per line has not # # been tested. It is possible that, if input is received in this form AND the # # newline characters have been truncated, the output could be garbled. # # # # --David Etlinger 6/7/2000 # # # # ADDED: Added a few lines of code to properly handle radio buttons. Checkboxes still need # # to be implemented. # # --David Etlinger 6/14/2000 # # # # ADDED: Added code to handle checkboxes. This is complicated because the submitted checkboxes # # are originally stored as a single string with "\0" as a delimiter. If the input type # # is determined to be checkboxes, the string is first split into an array. A hash key # # in a special checkbox array is then made to point to the array. This is done because # # there might be more than one checkbox set in a single question. Each time an input line # # of type checkbox appears, the next value in this array is popped into a temp variable. # # If it is determined that the line being processed corresponds to this value, the line # # is processed (made "sticky"); otherwise, the value is pushed back on the array. The # # fact that the number of checked cehckboxes is known but the total number of checkboxes # # is not means that a given line of input type checkbox might or might not correspond # # to the next value in the checkbox array. (I hope this explanation is clear enough!) # # --David Etlinger 6/28/2000 # ######################################################################################################### sub l2h_sticky_answers { my ( $rh_submittedAnswers, $ra_printLines, $rh_flags ) = @_; #warn ("rh_submittedAnswers = \@rh_submittedAnswers"); #warn ("ra_printLines = \@{ra_printLines}"); #warn ("rh_flags = \@{rh_flags}"); my %submittedAnswers = %{$rh_submittedAnswers}; my @printLines = @{$ra_printLines}; my @answerLabels = @{$rh_flags -> {ANSWER_ENTRY_ORDER}}; my $line; # holds the text of each line my $label; # holds each answer label my $counter = 0; # holds the index of the current answer my $output; # holds the text the subroutine returns my $answer_value; my %checkboxAns; # holder for the checkbox multi-part answers my $nextCheckboxAns; # temp holder for the next checkbox answer to be processed my $placeholder = "\x253"; # unused hex character to join text lines with #first, convert the array of text lines to one string... my $text = join( "$placeholder", @printLines ); #then, split it such that a line consists of either text #or a single <INPUT> tag (case insensitive; note also that #whitespace within the <INPUT> tag is accounted for). # NOTE -- the regular expression searches for "<", then any # amount of whitespace, then "INPUT", then any number of # characters that aren't ">", then ">". I think that instead of # searching for characters that aren't ">", I could have instead # searched to match a minimal number of characters (using ?), and # then ">". I don't know regular expressions well enough to tell # if this might lead to some subtle difference. my @textLines = split( m|(<\s*INPUT[^>]*>)|is, $text ); #my @textLines = split( m|(<\s*INPUT.*?>)|is, $text ); foreach $line ( @textLines ) { if( $line =~ m|<\s*INPUT|i ) { foreach $label ( @answerLabels ) { next unless exists( $submittedAnswers{$label} ); # skip if no answer was submitted. if( $line =~ m|NAME\s*=\s*"$label"|i ) { if( $line =~ m|TYPE\s*=\s*RADIO|i ) { #handle radio buttons $line =~ s|VALUE\s*=\s*"$submittedAnswers{$label}"|VALUE = "$submittedAnswers{$label}" CHECKED|i; } elsif( $line =~ m|TYPE\s*=\s*CHECKBOX|i ) { #make the hash key point to an anonymous array $checkboxAns{$label} = [ split( "\0", $submittedAnswers{$label} ) ] if not exists( $checkboxAns{$label} ); if( defined $checkboxAns{$label}[0] ) { $nextCheckboxAns = shift @{$checkboxAns{$label}}; if( $line !~ s|VALUE\s*=\s*"$nextCheckboxAns"|VALUE = "$nextCheckboxAns" CHECKED|i ) { unshift( @{$checkboxAns{$label}}, $nextCheckboxAns ); #put the unused answer back on the list } } } else { # we'll assume this is something else, like one or more fields. # if it's several fields, we need to take only one answer at a time # \0 are used to delimeter between entries. if ($submittedAnswers{$label} =~ /\0/ ) { my @answers = split("\0", $submittedAnswers{$label}); $answer_value = shift(@answers); # use up the first answer $submittedAnswers{$label}=join "\0", @answers; # store the rest $answer_value= '' unless defined($answer_value); } else { $answer_value = $submittedAnswers{$label}; } $line =~ s|VALUE\s*=\s*""|VALUE = "$answer_value"|i; } } } } #end if test for "<INPUT" $output .= $line; } #end foreach @printLines = split( m|$placeholder|, $output ); return @printLines; } #end l2h_sticky_answers() ## This is the old system (but newer than the one below). ## It has been replaced for two reasons: ## 1) It is complicated and difficult to understand or modify ## 2) It does not work for several situations that rarely come up, ## but must be handled properly. Specifically, it doesn't handle ## text with more than one <INPUT> tag on a given line very well. ## there are probably other problems, but that is the biggest. ## --DME 6/7/2000 # # the following doubly nested loop iterates over each line, # # and for each line searches for each answer label. Technically, # # it might have been faster to join each entry in @printlines # # into one string, search on that, and split it back up, but I # # felt that the slight theoretical speed gain was not worth the # # added complexity. # warn "answerLabels = @answerLabels"; #DEBUG # foreach $line ( @printLines ) { # warn "Line is $line"; #DEBUG # foreach $label ( @answerLabels ) { # if( $line =~ m|<INPUT TYPE=TEXT.*NAME="$label| ) { # while ($line =~ /VALUE = ""/) { # # Put trailing space in displayed answer so that while loop will # # always end. We are using the form of the s/// operator which # # evaluates its right hand side # $line =~ s|NAME="$label" VALUE = ""| # $counter++; # $submittedAnswers[$counter]=" " unless defined ($submittedAnswers[$counter]) # && not $submittedAnswers[$counter] =~ /^\s*$/; # qq{ NAME="$label" VALUE = "$submittedAnswers[$counter]" } |e; # # This insures that in VALUE = "$submittedAnswers[$counter]" # # the quantity $submittedAnswers[$counter] # # is never empty. This is required in order to terminate the loop. # } #end while # push( @output, $line ); # } #end if # else { # push( @output, $line ); # } # } #end foreach over @answerLabels # } #end foreach over @printLines # # @printLines = @output; # } #end outer if # # return @printLines; # } #end l2h_sticky_answers() ##subroutine that makes answers sticky in l2h mode # this is an old version of this routine, which assumes (incorrectly) # that answer labels begin with "AnSwEr". I've left it here just in case... # DME 6/6/2000 #sub l2h_sticky_answers { # my ($refSubmittedAnswers, $refprintlines)=@_; # my @printlines=@$refprintlines; # if ((@{$refSubmittedAnswers}!=0)) { # my $line; # my @output=(); # foreach $line (@printlines) { # if ($line =~ m|<INPUT TYPE=TEXT.*NAME="AnSwEr|) { # #print "<P>line doesn't exists<P>\n" unless defined($line); # while ($line =~ /VALUE = ""/) { # ## Put trailing space in displayed answer so that while loop will # ## always end. We are using the form of the s/// operator which evaluates its right hand side # $line =~ s|NAME="AnSwEr(\d*)" VALUE = ""| # my $tttemp = $1; # ${$refSubmittedAnswers}[$tttemp-1]=" " unless defined (${$refSubmittedAnswers}[$tttemp-1]) # && not ${$refSubmittedAnswers}[$tttemp-1] =~ /^\s*$/; # # qq{ NAME="AnSwEr$tttemp" VALUE = "${$refSubmittedAnswers}[$tttemp-1]" } |e; # # This insures that in VALUE = "${$refSubmittedAnswers}[$tttemp-1]" the quantity ${$refSubmittedAnswers}[$tttemp-1] # # is never empty. This is required in order to terminate the loop. # } # push(@output, $line); # } # else { # push(@output, $line); # } # } # # @printlines = @output; # } # # @printlines; #} ##subroutine that updates current keys in the l2h mode # sub l2h_update_keys { # my ($sessionKey, $refprintlines)= @_; # my @printlines=@$refprintlines; # my $line; # my @output=(); # #my $sessionKey = $main::sessionKey; # warn "hi lines = ",join("",@printlines); # foreach $line (@printlines) { # if ($line =~ m|^\s*<A(.*?)\&key=[^&]*&user|) { #<A.*&key=.*?&user # #grab the session key from the CGI input or make it blank # $line =~ s|^\s*<A(.*?)&key=[^&]*&user|<A$1&key=$sessionKey&user|; # warn "line = $line<BR>"; # push(@output, $line); # }else{ # push(@output, $line); # } # # } # @printlines; # # } sub makeL2H { my ($TMPPROBDIR,$psvn) =@_; $ENV{PATH} .= "$Global::extendedPath"; if($Global::externalLaTeX2HTMLVersion eq "98.1p1") { system("$Global::externalLaTeX2HTMLPath -no_math -init_file $Global::externalLaTeX2HTMLInit -dir $TMPPROBDIR -prefix $psvn $TMPPROBDIR${psvn}output.tex > $TMPPROBDIR${psvn}l2h.log 2>&1"); } elsif($Global::externalLaTeX2HTMLVersion eq "96.1") { system("$Global::externalLaTeX2HTMLPath -init_file $Global::externalLaTeX2HTMLInit -dir $TMPPROBDIR -prefix $psvn $TMPPROBDIR${psvn}output.tex > $TMPPROBDIR${psvn}l2h.log"); } else { die "Unknown LaTeX2HTML version: \$Global::externalLaTeX2HTMLVersion = $Global::externalLaTeX2HTMLVersion"; } } 1;