#!/usr/bin/perl

# processarchives.pl, copyright Casey Muller, let me know if you find
# it useful at processarchives(at)null-terminated.com and maybe I can
# give you a more up to date version.

# fourth beta version: Mar 1st, 2005

# TODO for fifth beta: fix RSS metadata

#directory constants
$webdir = `pwd`;
chomp $webdir;
$caseydir = $webdir."/root/casey0";
#$caseydir = $webdir."/staging/casey0";
$nulltdir = $webdir."/root/null-terminated";
#$nulltdir = $webdir."/staging/null-terminated";
$archivedir = $caseydir."/archive";
$tagdir = $caseydir."/tags/";

$imgprefix = "http://bigfiles.null-terminated.com:81/digitalcamera/";
$thumbpath = "/visuals/photography/by-date/";
$thumbprefix = "http://null-terminated.com".$thumbpath;

# structure: in each directory $archivedir/year/month there are
# ##.entry files which contain the entry for the ##th day. In them,
# the first line is the title, the second line is a timestamp (which
# is added if it doesn't exist), and the rest is stuck in the
# title.html file. Each index.html has links to all sub-titles (for
# now). The most recent 10 go in $webdir.

#load the feed templates
require "feedconstants.pl";

#use XML::RSS;
use POSIX;

#my $rss = new XML::RSS;
#$rss->channel(
#	      title       => $feedtitle,
#	      link        => $feedURL,
#	      description => $feeddesc,
#	      dc => {
#		  rights   => "Casey Muller under a Creative Commons license",
#		  language => "en-us",
#	      },
#	      syn => {
#		  updatePeriod    => "daily",
#		  updateFrequency => "2",
#		  updateBase      => "1980-06-24T08:00-08:00",
#	      },
#	      );

$today = strftime "%Y-%m-%dT%H:%M:%SZ", gmtime();
open(RSS, '>', "$caseydir/casey0.rss");
print RSS <<RSSHEADER;
<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title type="text">$feedtitle</title>
  <subtitle type="text">$feeddesc</subtitle>
  <updated>$today</updated>
  <link rel="alternate" type="text/html" href="$feedURL"/>
  <link rel="self" type="application/atom+xml" href="$feedRSSURL"/>
  <id>$feedRSSURL</id>
  <author>
    <name>Casey Muller</name>
    <email>atomfeed\@nullterminated.com</email>
  </author>
RSSHEADER


@months = qw(January February March April May June July August September October November December);
@monthnums{@months} = 1..$#months+1;

use File::Find;

# open the archive directory. .entry files get processed as entries

find { preprocess => sub {
    # sort by month slice if it's a year dir
    my $depth = (substr $File::Find::dir, length($archivedir)) =~ tr[/][];

    if($depth == 1) {
	sort {$monthnums{$a} <=> $monthnums{$b}} @_;
    } else {
	# need 2 to appear before 10
	sort {$a <=> $b} @_;
    }
}, wanted => sub {
    if(/(.*)\.entry(\d*)$/) {
	print "processing $_ in $File::Find::dir\n";
	
	# get just the archive part
	$dir = substr $File::Find::dir, length($archivedir);
	$archdir = substr $File::Find::dir, length($caseydir);

	# get date from filename and path
	$day = $1;
	
	# split it up
	@tracedirs = split m%/%, $dir;
	shift @tracedirs;
	$year = shift @tracedirs;
	$month = shift @tracedirs;
	$monthnum = sprintf "%02s", $monthnums{$month};
	$daynum = sprintf "%02s", $day;
	
	# open it
	open(INPUT, "$File::Find::name") 
	    || die "can't read $File::Find::name";
	
	$title = <INPUT>;
	chomp $title;

	$tagline = <INPUT>;
	chomp $tagline;
	@tags = split ' ', $tagline;

	# what's the relative url?
	$url = "$archdir/$title.html";
	$url =~ y/ /_/;

	# save info to add to rss and index
	$indexhtml = "<a class=\"title\" href=\"$url\">$title</a> <div class=\"date\"><a href=\"$url\">$month $day, $year</a></div>\n";

	unshift @allindex, $indexhtml;
	unshift @yearindex, $indexhtml;
	unshift @monthindex, $indexhtml;

	unshift @rsstitle, $title;
	unshift @rssurl, "$feedURL$url";
	unshift @rssdate, "$year-$monthnum-$daynum"."T12:00:00-08:00";

	# open the output file
	open(OUTPUT, '>', "$caseydir$url") 
	    || die "can't write $caseydir$url";
 
	# write out the headers
	print OUTPUT $pretitle;
	print OUTPUT $title;
	print OUTPUT $title2date;
	print OUTPUT "$month $day, $year";
	print OUTPUT $date2content;
	
	# dump the actual file:
	$entry = "";
	$rssentry = "";
	$#rsstags = -1;

	foreach $tag (@tags) {
	    $tag = lc $tag;
	    print "TAG: $tag\n";
	    # do we need a new file
	    if(!exists $tagfiles{$tag}) {
		print "creating file for tag: $tag\n";
		open($tagfiles{$tag}, '>', $tagdir.$tag.".html");
		print { $tagfiles{$tag} } <<HTMLBLOCK;
<div class="tagheader"><h1><a href="index.html">tag</a>: $tag</h1></div>
HTMLBLOCK
            }
	    # add ourselves
	    print { $tagfiles{$tag} } $indexhtml;

	    $entry .= "<a class=\"tag\" href=\"/tags/$tag.html\" rel=\"tag\">tag: $tag</a>";
	    $rssentry .= " <a href=\"$feedURL/tags/$tag.html\" rel=\"tag\">tag: $tag</a>";
	    push @rsstags, $tag;

	    # increment the counters
	    $tagcount{$tag}++;
	}

	$entry .= "\n";
	$rssentry .= "\n";
	while(<INPUT>) {
	    # duplicated code:
	    s#<limg: *([^ ]+)/([^ ]+) (.+)>#<a name=\"$1/$2 $3\" href=\"$thumbprefix$1/$2.html\"><img style=\"float: left; clear: left; margin: 4px;\" src=\"$thumbprefix$1/thumb$2\"/>$3</a>#g;
	    s#<rimg: *([^ ]+)/([^ ]+) (.+)>#<a name=\"$1/$2 $3\" href=\"$thumbprefix$1/$2.html\"><img style=\"float: right; clear: right; margin: 4px;\" src=\"$thumbprefix$1/thumb$2\"/>$3</a>#g;
	    if(/^$/) {
		if($rssentry) {
		    # if it's a non-leading newline, use <br />
		    $rssentry .= "  <br />\n<br />\n";
		    $entry .= "  </p>\n  <p>\n";
		}
	    } else {
		$entry .= $_;
		$rssentry .= $_;
	    }
	}
	close INPUT;
	# strip trailing newlines
	$entry =~ s/[\s\n]+$//;
	# write it to the actual page
	print OUTPUT $entry;
	if($#allindex > 0) {
	    print OUTPUT $content2prev;
	    print OUTPUT $archiveentrystart;
	    print OUTPUT $allindex[1];
	    print OUTPUT $archiveentryfinish;
	}

	print OUTPUT $postprev;
	close OUTPUT;

	# save a copy for the main page and truncate for rss
	unshift @entries, $entry;
	unshift @rssdesc, $rssentry;
	unshift @rsssubjects, [ @rsstags ];

	return;
    }
}, postprocess => sub { 
    print "creating index for $File::Find::dir\n";
    
    # open the output file
    open(OUTPUT, '>', "$File::Find::dir/index.html") 
	|| die "can't write $File::Find::dir/index.html";

    # are we a month or a year?
    $dirs = (substr $File::Find::dir, length($archivedir)) =~ tr[/][];
    print "got $dirs directories\n";
    if($dirs == 0) {
	# the main archive page has all entries
	print OUTPUT $archivestart;
	foreach (@allindex) {
	    print OUTPUT $archiveentrystart;
	    print OUTPUT $_;
	    print OUTPUT $archiveentryfinish;
	}
	print OUTPUT $archivefinish;

	# also create the frontpage
	open(MAIN, '>', "$caseydir/index.html") 
	    || die "can't write $caseydir/index.html";
	print MAIN $mainstart;
	# add the recent photos at the top
        open(RECENTPHOTOS, "$webdir/recentPhotos");
        while(<RECENTPHOTOS>) {
	    print MAIN $_;
	}
	close RECENTPHOTOS;
	# assume sorted by date, so take the last 5 for the main page/rss
 	if($#allindex>=5) {
 	    @allindex = splice(@allindex, 0, 5);
 	    @entries = splice(@entries, 0, 5);
 	    @rsstitle = splice(@rsstitle, 0, 5);
 	    @rssurl = splice(@rssurl, 0, 5);
 	    @rssdesc = splice(@rssdesc, 0, 5);
 	    @rssdate = splice(@rssdate, 0, 5);
 	    @rsssubjects = splice(@rsssubjects, 0, 5);
 	}

	open(NULLTINDEX, '>', $nulltdir."/index.html");
	print NULLTINDEX "The latest entry from my <a href=\"http://casey0.com/\">project log</a> is titled <a href=\"$rssurl[0]\">$rsstitle[0]</a>. It was posted on <a href=\"$rssurl[0]\">$rssdate[0]</a>.\n\nSee more <a href=\"/self\">about all this</a> or just start browsing the categories.\n\nRecent photographs and projects:<br/>";
	open(RECENTPROJS, $webdir."/recent")
	    || die "can't read $File::Find::name";
	while(<RECENTPROJS>) {
	    print NULLTINDEX $_;
	}
	close RECENTPROJS;
	close NULLTINDEX;

	while($#allindex+1) {
	    print MAIN $mainprecontent;
	    print MAIN shift @allindex;
	    print MAIN shift @entries;
	    print MAIN $mainpostcontent;

	    $tmptitle = shift @rsstitle;
	    $tmplink = shift @rssurl;
	    $tmpdesc = shift @rssdesc;
            $tmpdesc =~ s/\&/\&amp\;/g;
            $tmpdesc =~ s/\</\&lt\;/g;
            $tmpdesc =~ s/\>/\&gt\;/g;
	    $tmpdate = shift @rssdate;
	    @tmpcats = @{ shift @rsssubjects };
	    $tmpcat = "";
	    foreach $cat (@tmpcats) {
		$tmpcat .= "<category term=\"$cat\"/>";
	    }

#			       subject => join(' ', @{ shift @rsssubjects }),
	    print RSS <<RSSBLOCK;
  <entry>
    <title>$tmptitle</title>
    <link rel="alternate" type="text/html" href="$tmplink"/>
    <id>$tmplink</id>
    <published>$tmpdate</published>
    <updated>$tmpdate</updated>
    <summary type="html">$tmpdesc</summary>
    $tmpcat
  </entry>
RSSBLOCK
#	    $rss->add_item(
#			   title       => shift @rsstitle,
#			   link        => shift @rssurl,
#			   description => shift @rssdesc,
#			   dc => {
#			       date => shift @rssdate,
#			       subject => join(' ', @{ shift @rsssubjects }),
#			      }
#			   );
	}
	print MAIN $mainfinish;
	close MAIN;
    } else {
	if($dirs == 1) {
	    print OUTPUT $yearormontharchive;
	    foreach (@yearindex) {
		print OUTPUT $archiveentrystart;
		print OUTPUT $_;
		print OUTPUT $archiveentryfinish;
	    }
	    $#yearindex = -1;
	} elsif($dirs == 2) {
	    print OUTPUT $yearormontharchive;
	    foreach (@monthindex) {
		print OUTPUT $archiveentrystart;
		print OUTPUT $_;
		print OUTPUT $archiveentryfinish;
	    }
	    $#monthindex = -1;
	}
    }

    close OUTPUT;

}}, $archivedir;

# save rss file
#$rss->save("$caseydir/casey0.rss");

print RSS "</feed>\n";
close RSS;

# fix rss file, allow multiple subjects
#use Tie::File;
#tie @rsslines, 'Tie::File', "$caseydir/casey0.rss";
#foreach (@rsslines) {
#    if(/\<dc:subject\>(.+)\<\/dc:subject\>/) {
#	@tags = split / /, $1;
#	$_ = "";
#	foreach $tag (@tags) {
#	    $_ .= "<dc:subject>$tag</dc:subject>"
#	}
#    }
#}

open(TAGS, '>', $tagdir."index.html");
print TAGS "All tags:\n";

foreach(keys %tagfiles) {
    if($tagcount{$_} > $maxtagcount) {
	$maxtagcount = $tagcount{$_};
    }
    if($tagcount{$_} < $mintagcount) {
	$mintagcount = $tagcount{$_};
    }
}

foreach(sort keys %tagfiles) {
    # interpolate between 12 and 36
    $fontsize = ($tagcount{$_} - $mintagcount) / $maxtagcount * 24 + 12;
    print TAGS "<a href=\"$_.html\" style=\"font-size: ".$fontsize."px;\">$_</a>($tagcount{$_}) \n";
    print "$_ = $tagcount{$_}\n";
    close $tagfiles{$_};
}

close TAGS;