# For the converting of Blogger archives into cleaner MT import files.
#
# A script by sysadmin1138
#
# This script expires March 26th 2010, when Blogger terminates 
# the ability to use FTP to publish. At that point, wget/curl 
# will be needed to retrieve the files needed. You're on your
# own for that.
#
# The Blogger template needs to be changed to the following lines:
#
# <blogger><ArchivePage>
# EAUTHOR:::yourmtusername
# ETITLE:::<PostSubject><$BlogItemSubject$></PostSubject>
# EDATE:::<$BlogItemDateTime$>
# EBODY:::<$BlogItemBody$>
# <BlogItemComments>
# CAUTHOR:::<$BlogCommentAuthor$>
# CDATE:::<$BlogCommentDateTime$>
# CBODY:::<$BlogCommentBody$>
# </blogitemcomments>
# </ArchivePage></Blogger>
#
# This sets up the files the way the script expects.
# Blogger Timestamps need to be in MM/DD/YYYY HH:MM AM/PM mode.

use strict;

my $SectionMarker="-----";
my $EntryMarker="\n--------";
my $InName;
my $OutName;
local *InFile;
local *OutFile;
my $EAuthor;
my $CAuthor;
my $AIndex;
my $ATemp;
my $Title;
my $EDate;
my $CDate;
my $PCategory;
my @Category;
my $Label;
my $Labels;
my $LabelIndex;
my $EndLabelIndex;
my $LabelSize;
my $CUrl;
my $EBody;
my $CBody;
my $PLine;
my @Split;
my $spat=":::";
my $EntryCount=0;

# Provide usage information
if ($#ARGV != 1) {
    print "\nUsage:\tblogger2mt.pl <blogger-filename> <outputname>\n\nExample:\n\tblogger2mt.pl 2001_10_01_archive.txt ~/www/cgi-bin/mt/import/2001_10_01_mtimport.txt\n\n";
    die;
}

# Open the files we'll be working with
$InName=$ARGV[0];
$OutName=$ARGV[1];
open (InFile, "<:utf8", $InName);
open (OutFile, ">:utf8", $OutName);


# Start processing the file!
$PLine=readline(InFile);
while (defined($PLine)) {
    chomp($PLine);
    @Split=split(/:::/, $PLine);
    if ($Split[0] eq "EAUTHOR") {                     # This is the first line of a section, so clear variables
        undef($EAuthor);
        undef($EDate);
        undef($Title);
        undef($EBody);
        undef(@Category);
        undef($PCategory);
        $EAuthor=$Split[1];
        $EntryCount++;
    } elsif ($Split[0] eq "ETITLE") {
        $Title=$Split[1];
    } elsif ($Split[0] eq "EDATE") {
        $EDate=$Split[1];
    } elsif ($Split[0] eq "CAUTHOR") {
        undef($CUrl);
        undef($CAuthor);
        #Trim the terminal ">" of the Author line
        chop($Split[1]);
        $AIndex=rindex($Split[1], '>');
        $CAuthor=substr($Split[1], $AIndex+1);
        $AIndex=index($CAuthor, '<');
        $CAuthor=substr($CAuthor, 0, $AIndex);
        if ($Split[1] =~ /href/) {
            $AIndex=index($Split[1], 'href');
            $CUrl=substr($Split[1], $AIndex+6);
            $AIndex=index($CUrl, '"');
            $CUrl=substr($CUrl, 0, $AIndex);
        }
    } elsif ($Split[0] eq "CDATE") {
        $CDate=$Split[1];
    } elsif ($Split[0] eq "CBODY") {
        $CBody=$Split[1];
        # Since the Body is the last thing in the comment, commit the comment-block
        print (OutFile "\nCOMMENT:\n");
        print (OutFile "AUTHOR: $CAuthor\n");
        if (defined($CUrl)) {
            print (OutFile "URL: $CUrl\n")
        }
        print (OutFile "DATE: $CDate\n");
        print (OutFile "$CBody\n");
        print (OutFile "$SectionMarker\n");
    } elsif ($Split[0] eq "EBODY") {
        # First, extract the body of the post from any attached labels
        $LabelIndex=index($Split[1], '<p class="blogger-labels">');
        if ($LabelIndex != -1) {
            $EBody=substr($Split[1], 0, $LabelIndex);
        } else {
            $EBody=$Split[1];
        }
        # Then parse any labels, turning them into categories
        $Labels=substr($Split[1], $LabelIndex);
        $LabelIndex=index($Labels, 'txt">');
        while ($LabelIndex != -1) {
            $EndLabelIndex=index($Labels, '</a>', $LabelIndex);
            $LabelSize=$EndLabelIndex - $LabelIndex;
            $Label=substr($Labels, $LabelIndex+5, $LabelSize-5);
            $LabelSize=push(@Category, $Label);
            if ($#Category == 0 ) {
                $PCategory=$Label;
            }
            $Labels=substr($Labels, $EndLabelIndex);
            $LabelIndex=index($Labels, 'txt">');
        }
        # Finally, output the post so far. Comments will be handled separately.
        print (OutFile "$EntryMarker\nAUTHOR: $EAuthor\n");
        print (OutFile "TITLE: $Title\n");
        print (OutFile "STATUS: Draft\n");
        if ($#Category > -1) {
            print (OutFile "PRIMARY CATEGORY: $PCategory\n");
            for (my $i=0; $i <= $#Category; $i++) {
                print (OutFile "CATEGORY: $Category[$i]\n");
            }
        }       
        print (OutFile "DATE: $EDate\n$SectionMarker\n");
        print (OutFile "BODY:\n$EBody\n");        
        print (OutFile "$SectionMarker\n");
        print (OutFile "EXTENDED BODY:\n\n$SectionMarker\n");
        print (OutFile "EXCERPT:\n\n$SectionMarker\n");
        print (OutFile "KEYWORDS:\n\n$SectionMarker\n");
    }
    $PLine=readline(InFile);
}

# Finalize the file
print (OutFile "\n$EntryMarker\n");
# Close up files
close InFile;
close OutFile;
print "Processed $EntryCount entries in file $InName.\n";