#!/usr/bin/env perl ## $Id$ #################################################################### # Copyright @ 1995-1998 University of Rochester # All Rights Reserved #################################################################### use strict; use lib '.'; use webworkInit; # WeBWorKInitLine # this is the home directory. (change to system) use Global; use Pod::Html; use Carp; use CGI qw(:standard); #eval { $| = 1; #force command buffering print "Content-type: text/html\n\n"; $| = 0; # augment the paths to search for documentation. # e.g. '/webwork/experimental/courseScripts', '/webwork/experimental/scripts', # This is not fool proof, but it will work if the webwork system has been set up as at U. of Rochester. #my $dir3 = "/ww/webwork/development/"; #kludge for the summer so that the latest docs make it to the web. my $dir3 = $Global::mainDirectory; unshift(@INC,$dir3); unshift(@INC, $dir3.'courseScripts/',$dir3.'scripts/',$dir3.'cgi/cgi-scripts/'); my $dir4 = "/u/gage/webwork-modperl/lib/"; unshift( @INC, $dir4.'Apache/', $dir4.'WeBWorK/', $dir4.'WeBWorK/PG/', $dir4.'WeBWorK/ContentGenerator/' ); # End defining search paths. open( LOG, ">pod2webwork.log" ) or warn "Can't open pod2webwork.log for writing"; #print LOG ""; #close( LOG ); #open( LOG, ">>pod2webwork.log" ); my $path = param('path'); my $HTMLpre = 0; #for use with clipline(); must be global to this file #because it's used over multiple calls of clipline() my $up_path = referer(); #This is the URL that the "up" button will point to. if (defined $up_path) { # point to the index of the directory $up_path =~ s|/[^/]*$|/index.html|g; } else { $up_path = "${Global::webworkDocsURL}"; } # Now find the path to the file my $p; foreach $p ("",@INC) { if (-r "$p$path") { $path = "$p$path"; #print "found $path
\n"; last; } #print "looking for $p$path
\n"; } output_filter(); # install output filter pod2html($path, "--title= "); # produces a blank title, eliminating error message. # real title is produced in output_filter close(STDOUT); close( LOG ); #}; # end eval print "Content-type: text/html\n\n 
 $@
" if $@; ########### # This will print to STDOUT which will then be filtered. sub output_filter{ # use command buffering; my $pid; return if $pid = open(STDOUT, "|-") ; #I'm not sure why this works -- see Perl cookbook. print < $path

WeBWorK

[Up] [Feedback]

EOT unless (-r $path) { print "

Missing documentation

"; print "Can't find file $path after searching in the following directories:
"; print join("
",@INC); } unless ( defined($pid) ) { print "

Can't find the file $path

"; return; } # get rid of the html header while () { last if //; } while () { my $line = $_; last if $line =~ m{}i; chomp($line); print clipline($line); } print <
File path = $path
Last updated: 12 August 1999  $path
WeBWorK documentation written by M. Gage, E. Pachepsky and A. Pizer.
This page was built by M.Gage EOT exit; } # end output filter # Routine authors: Arnold Pizer and Michael Gage # modified by David Etlinger # # This routine handles indenting the line and wrapping it, and other formatting # issues. The incoming text includes HTML tags; code which wraps lines at a # certain number of characters must ignore these tags. The assumptions I'm making # about the incoming text are: # 1) It contains HTML tags already # 2) Special characters (like &) are replaced by HTML codes (like &) # 3) Most tabs have already been replaced by spaces (this is a pain to deal with!) # Pod2HTML uses 8-column tabs, while we want to use 4-column. Also, 2 or # more consecutive spaces are assumed to be a tab. So incoming 8-column # tabs need to be converted to 4-column tabs. # Some of these assumptions might not be valid if pod2HTML changes, or other # system components are modified. # --David Etlinger 7/27/2000 sub clipline { my $inLine = shift @_; #DEBUG # $inLine =~s/"([^;])/"$1/g; # work around for bad quoting in pod2html for links print LOG "${inLine}\n"; $inLine = '' unless $inLine =~ /\S/; #treat lines with only whitespace as blank $HTMLpre = 1 if $inLine =~ m|
|i;		#$HTMLpre assumes 
 begins the line
	
	$inLine =~ s/``/"/g;		#use " for increased browser compatibility
	$inLine =~ s/''/"/g;
	# $inLine =~ s/"/"/g;    # can't do this you destroy legitimate HTML attributes
	my $p2Htabs = 8;				#pod2HTML uses 8-column tabs
	
	#Replace multiple spaces with tabs. $& is the string that was matched.
	#The number of tabs is equal to the number of spaces / $p2Htabs, plus 1 if there
	#is a remainder, 0 if not. See comments above for tab explanation.
	1 while $inLine =~
		s|  +|"\t" x ( int(length($&) / $p2Htabs)+(length($&) % $p2Htabs != 0 ? 1 : 0) )|e;
		
		
	# kludge to work around the fact that sometimes pod2html uses " instead of "

	my @line = split( //, $inLine );
	
	# constants
	my $tabWidth = 4; 
	my $indentFlag = 1;
	my $commentIndentInc = 2;  # zero puts t
	my $maxLineLength = 80;
	
	# counters
	my $HTMLtag = 0;		#is the current character part of an HTML tag?
	my $HTMLchar = 0;		#are we in special character of the form &name; ?
	my $indentSize = 0;
	my $position = 0;		#give number of characters to the left of the pointer
							#	(pointer is after current character)
	my $spacesForTab = 0;	#number of spaces to substitute for a tab

	my $out = "";
	
	foreach my $char ( @line ) {
		$position++;
		
		if( $char eq '<' ) {		#entering an HTML tag
			$HTMLtag = 1;
		}
		if( $char eq '&' ) {		#entering HTML special character
			$HTMLchar = 1;
		}
		
		$position -= ($HTMLtag || $HTMLchar);	#don't count HTML tags or chars for
												#purposes of line breaks and tabs

		if( $char =~ /\s/ ) {
			die "clipline(): The line cannot contain a return" if $char eq "\n";
           
			if( $char eq "\t" ) {
				#emulates tab stops -- the tab might or might not be the full
				#width of the tab field, depending on where it started
				$spacesForTab = ( ($position % $tabWidth) == 0 ?
										1 : ($tabWidth - $position % $tabWidth + 1) );
				
				$char = ' ' x $spacesForTab;
				
				$indentSize += $spacesForTab if $indentFlag == 1;
				$position += $spacesForTab - 1;
			}
			else {
				$indentSize++ if $indentFlag == 1;
			}

#Uncomment to enable wrapping lines after $maxLineLength characters.
#This code might need updating to work correctly
			if ($position > $maxLineLength) {
				$out .= $char . "\n" . ' ' x $indentSize;
				$position = $indentSize + 1;
			}
		}
		else {
			$indentFlag = 0 if $HTMLtag == 0;
		}
						
		if ($char eq "#") {
			$indentSize = $position + $commentIndentInc - 2;  #the 2 is right
		}
		### debug  $char = "$char|$position|" if $char eq "|";
		$out .= $char;
		
		if( $char eq '>' && $HTMLtag == 1 ) {		#leaving an HTML tag
			$HTMLtag = 0;
		}
		if( $char eq ';' && $HTMLchar == 1 ) {		#leaving an HTML char
			$HTMLchar = 0;
			$position++;
		}
	}
	
	$out =~ s/\s*$//;
	
	#this was an attempt to force line breaks to follow the source text, but
	#incoming text has already removed those line breaks.	DME 8/1/2000
	#
	#don't add a 
 if the line is blank or contains only HTML tags
	#$out .= "
" if( $HTMLpre == 0 && $out ne '' && $out !~ m|^\s*\s*$| );
	
	$out .= "\n";
	
	$HTMLpre = 0 if $out =~ m|
|i;		#assumes 
 ends the line
	
	return $out;
}