#!/usr/bin/perl -w
# eozl's alml2html convertor and consolidator. ask him about alml.
# used in my online journals. proto-megamesh.
# written aug 13-18, 2004
# ignore copyrights.

use warnings;
use CGI;
use Time::HiRes qw(gettimeofday tv_interval );
use POSIX;


$time0 = [gettimeofday]; #set up initial time
print "Content-type: text/html\n\n";

# these ENV vars are useful:
#[DOCUMENT_URI] -> [/html/evilhaxx0ring/perl.html]
#[DOCUMENT_ROOT] -> [/var/www/localhost/htdocs]

#find the directory of the html
$entrydir = $ENV{DOCUMENT_ROOT} . substr($ENV{DOCUMENT_URI}, 0, rindex($ENV{DOCUMENT_URI}, '/')) . '/entries';


#read all the files there, push into @entryfiles
opendir(DIR, $entrydir) or die "died trying to open $entrydir. $!";
@entryfiles=();
while (defined ($_ = $entry = readdir(DIR))) {
    push(@entryfiles,"$entrydir/$entry") if (-T "$entrydir/$entry" && /\.entry$/); 
}

@entryfiles=(reverse sort @entryfiles);



# process all the files
# set up each entry header, css, and alml2html it.
foreach $entry (@entryfiles) {
    open (DATA, $entry) or die "died while reading $file; $!";
    $str = do { local $/, <DATA> };
    close DATA;
    $str =~ s/\s+$//; #clear whitespace at the end of the entry

    alml2html($str);  #convert the entry to alml

    my ($dir, $f) = $entry =~ m/(.*\/)(.*)$/ ;
    ($title) = $f =~ m/(.*).entry/;
    #find the last modified timestamp
    my $mtime=(stat $entry)[9]; 
    $mtimestr = lc(strftime("%a %b %d %H:%M:%S %Y", localtime($mtime)));
    
    print <<EOHTML;
    <p><a name="$title"> </a>\n<span class="date"><a href="#$title">$f</a> | <a href="entries/$f">$mtimestr</a></span>
    <div class="entry">
EOHTML
    print "$str</div>\n\n\n";

}
$elapsed = tv_interval ($time0);
print "<!-- time to do: $elapsed -->";

exit;




###
# # ALML-2-HTML stuff below
###

sub almlital {
    $str =~ s/\.\\(.*?)\.\\/<i>$1<\/i>/g;
}

sub almllink {
# NOTE:
## </link> convertor for <> is not here. put this after <> is done.
## $str =~ s/<\/link>/<\/a>/g; # to take care of </link> used in <>
    ### <link ./local.html> <link /local>
    ### <link xtp://external.com>
    ### <link packages.gentoo.org>
    $str =~ s/<link (\"?)([.\/].*?)\1 ?>/<a href=\"${2}\">/g; # /local ./local
    $str =~ s/<link (\"?)([^>]*?tp:\/\/.*?)\1 ?>/<a  href=\"${2}\">/g;  #xtp://extern
    $str =~ s/<link (\"?)(.*?)\1 ?>/<a   href=\"http:\/\/${2}\">/g; #extern.com

}

sub almlparen {
    #use temporary string for escaped \(s ... 
    #if there's a better way, tell me.
    $str =~ s/\\\(/_\(_/g;
    $str =~ s/\\\)/_\)_/g;

    $str =~ s/\(/<font size="-1">\(/g;
    $str =~ s/\)/\)<\/font>/g;

    $str =~ s/_\(_/\(/g;
    $str =~ s/_\)_/\)/g;

    #$str =~ s/\\\(/\(/g; #screws up (())
    #$str =~ s/^\(/<font size="-1" >\(/;  #for initial (s
    
    #$str =~ s/([^\\])\)/${1}\)<\/font>/g; #screws up (())
    #$str =~ s/\\\)/\)/g;screws up (())
}


sub alml2html {
    ### remaining scutwork

    #initial PRE
    if (0) {
    @precontents = ($str =~ m/<pre>(.*?)<\/pre>/gs);
    $prei=0;
    while ( $str =~ s/<pre>(.*?)<\/pre>/__pre${prei}__/s) {
	$prei++;
    }}


    # deal with annoying .\italics.\
    almlital($str);
 
    # link: note </link> convertor is at the end here
    almllink($str);
   
    # deal with ()s parens
    almlparen($str);


    #deal with old <l> shortform of <li>...
    $str =~ s/<l>/<li>/g;

    #deal with \>-)
    $str =~ s/\\</&lt;/g;
    $str =~ s/\\>/&gt;/g;
    

    ###    
#####deal with "quotes": set up array of " and <tag> positions
    ###

    #$str =~ s/\"(\w*)\"/~${1}~/g;
    $str =~ s/\\\"/&quot;/g;
    
    ## make quotes list
    @quots=();
    $n = -1;
    while (1) {
	$_=index($str, "\"",$n);
	if ($_ == -1) {
	    last;}
	push(@quots, $_);
    $n=1+$_;
    }
    ## make tags list
    @tags=();
    $n = -1;
    while (1) {
	$lt = index($str, "<", $n);
	$gt = index($str, ">", $n);
	#print "lt gt: $lt $gt\n";
	if (($lt == -1) && ($gt == -1)) {last;}
	push(@tags, $lt, $gt );
	#if we ever expect this to be needed...
	##if ($lt < $gt) { $n = $gt +1 ; } else {$n = $lt +1 ;}
	$n = $gt + 1;
    }

    ## purge quotes inside tags
    $t = 0;
    $q = 0;
    @tagclose=();
    while ($t <= $#tags) { #not foreached because we need to increm by 2

	#set up @tagclose for <>
	if ($tags[$t] +1 == $tags[$t+1]) { 
	    $prevtag=substr($str, $tags[$t-2], 1+$tags[$t-1]-$tags[$t-2]);
	    #print "phoobe: $t, $tags[$t-1], $prevtag \n<br>";
	    $prevtag =~ s/<(.*?)(?:( .*?)>|(?)>)/<\/${1}>/g;
	    
	    push(@tagclose, $prevtag);
	    $t+=2;
	    next;
	}
	foreach $q (0..$#quots) {
	    #while ($q <= $#quots) {
	    if  (($quots[$q] > $tags[$t]) && ($quots[$q] < $tags[$t+1])){
		#delete $quots[$q];
		$quots[$q]=-1;
	    }
	}
	$t+=2;
    }
    
    ## rebuild the string
    $i=0;
    $qflag = 0;
    $offsets=0;
    foreach $i (0..$#quots) {
	if ($quots[$i] == -1) {
	    next;
	}
	if ($qflag == 0) {
	    substr($str, $quots[$i]+$offsets, 1) = "<tt>\"";
	    $offsets+=4;
	    $qflag = 1;
	}
	else {
	    substr($str, $quots[$i]+$offsets, 1) = "\"</tt>";
	    $offsets+=5;
	    $qflag = 0;
	}
    }
    # take care of <>'s
    $n = -1;
    $t=0;
    while (1) {
	$_=index($str, "<>",$n);
	if ($_ == -1) {
	    last;}
	substr($str, $_, 2) = $tagclose[$t];
	$t++;
	$n=1+$_;
    }
    #
#####FINISH QUOTES STUFF
    #

    # to take care of </link> used in <>
    $str =~ s/<\/link>/<\/a>/g;

    #deal with \n\n-><p>s parags, \n-><br> breaks
    $str =~ s/\n\n/<p>/g;
    $str =~ s/\n([^ ])/\n<br>$1/g;


    # Remove <br> and <p> from <pre> things
    while (($breaker) = $str =~ m/<pre>.*?(<br>|<p>).*?<\/pre>/gs) {
	$str =~ s/<pre>(.*?)${breaker}(.*?)<\/pre>/<pre>${1}${2}<\/pre>/gs;
    }

    # NOT USED
    #another method of removing all tags from PREs...
    if (0) {
    #final PRE
    foreach $prei (0..$#precontents) {
	#print "prei: $prei->$stf[$prei]\n";
	$prestr="__pre${prei}__";
	$str =~ s/$prestr/<pre>@precontents[$prei]<\/pre>/gs;
    }}
    # i r teh jeo!!!

}