#!/usr/bin/perl

use strict;
use warnings;
use diagnostics;

use File::Find;
use Date::Format;
use Date::Parse;

sub process {
    if (-f $_ && /\.html$/) {
        my $fn = $_;

        open(F, "<:utf8", $fn) or die("Cannot open ${File::Find::name}: $!");
        my $file = <F>;
        close F;

        my ($text, $time);

        ($text = $1) =~ s/\r\n//mg if ($file =~ m%<h3 class="post-title">(.*?)</h3>%s);
        $text .= $1 if ($file =~ m%<div class="post-body">(.*?)</div>%s);
        $time = str2time($1) if ($file =~ /\w+, (\w+ \d\d, \d{4})/s);

        die "Invalid fileformat: $text $time." unless ($text && $time);

        $text =~ s/(\r|<p>|<\/p>|^\s*|[\t ]{2,})//g;
        $fn =~ s/html$/txt/;

        open(OUT, ">:encoding(iso-8859-15)", "/home/hynek/old-blog/new/$fn") or die "Cannot open for write: $!.";
        print OUT $text;
        close(OUT);
        utime $time, $time, ("/home/hynek/old-blog/new/$fn");
    }
}

undef $/;
find(\&process, ("tmp"));