foswiki2mediawiki

#!/usr/bin/perl
#
# foswiki2mediawiki converts FOSWiki sourcecode to MediaWiki
#
#   The MediaWiki importTextFile tool would then import the file into mediawiki
#
# Derived from:
#   https://bl0rg.net/software/twiki2mediawiki/
#
#  2014-05-01 - ASM
#
#   Changed script to simply convert one file from FOSwiki to MediaWiki syntax
#     and write the file to the same directory. 
#
#
# BUGS 
# - Leaves some HTML-markup that worked in FOSWiki hanging around
# - Doesn't handle *fat* FOSWiki-style
# - Doesn't handle Inter-TwikiWeb.Links correctly
# - Doesn't handle all styles of links 
# 
# - Rename a lot of pages (CameLcase)
# - Fix b0rken links (see BUGS)
# - Fix b0rken markup (see BUGS)
#
#
# TODO :
#  - Handle attachments better
#
#

use strict; 
use Getopt::Std;

# temp
use Data::Dumper;


# Options
my %opts;
getopts('f:d', \%opts);


sub usage {
  print "\n";
  print "  Usage:   $0 [-d] -f filename \n";
  print "\n";
  exit;
}


main();

sub main {

  # take  filename from -f in getopts
  (!exists $opts{f}) && usage() ;

  my $foswikifile = $opts{f};
  (!-e $foswikifile ) && die "File: $foswikifile  does not exist.\n"; 

  # maybe check for .txt extension  and remove to form output filename

  # for now just output to STDOUT

  # file handle
  my $fh;
  open(my $fh, "<", "$foswikifile") or die "cannot open < $foswikifile : $!";

  # INPUT
  my $input;
  while  (<$fh>) {
    $input .= $_;
  }


  # 
  # Show input if debug flag is set
  #
  if ( $opts{d} ) {
    print "--------------------------------------------\n";
    print "INPUT: \n";
    print $input;
    print "\n";
    print "--------------------------------------------\n";
    print "OUTPUT: \n";
  }

  my $output = convert_from_foswiki($input);
  print $output;
  print "\n";
  
  if ( $opts{d} ) {
    print "--------------------------------------------\n";
  }

  exit;
}




sub convert_from_foswiki {
  my $cur_text = shift;

  # HTML comments are lame and used by Twiki for stuff
  $cur_text =~ s#<!--.*?-->##gms;
  
  # Handle special fields
  $cur_text =~ s/%META:FIELD{(.*?)}%/$1/g;

  # Strip FOSWiki meta-tags
  $cur_text =~ s/%.*?%//g;

  # strip ! preceding a (wiki)word
  $cur_text =~ s/ !/ /gm;

  # Convert headings
  #  add good num of leading newlines
  # h4-h7 get converted to h4
  $cur_text =~ s/\n*---\+\+\++\s*(.*)/\n\n==== $1 ====/gm;
  $cur_text =~ s/\n*---\+\+\++\s*(.*)/\n\n=== $1 ===/gm;
  $cur_text =~ s/\n*---\+\+\s*(.*)/\n\n\n== $1 ==/gm;
  # MediaWiki style generally does not use H1 tags, just bold it. 
  $cur_text =~ s/\n*---\+\s*(.*)/\n\'\'\'$1\'\'\'/gm;

  # Convert tabbed lists to "*"-lists
  $cur_text =~ s/^\t\*/\*/gm;
  $cur_text =~ s/^\t\t\*/\*\*/gm;
  $cur_text =~ s/^\t\t\t\*/\*\*\*/gm;
  $cur_text =~ s/^\t1/#/gm;
  $cur_text =~ s/^\t\t1/##/gm;
  $cur_text =~ s/^\t\t\t1/###/gm;

  # Convert 3-space lists to "*"-lists
  $cur_text =~ s/^   \*/\*/gm;
  $cur_text =~ s/^      \*/\*\*/gm;
  $cur_text =~ s/^         \*/\*\*\*/gm;
  $cur_text =~ s/^   1/\#/gm;
  $cur_text =~ s/^      1/\#\#/gm;
  $cur_text =~ s/^         1/\#\#\#/gm;

  # Convert umlauts and other funny characters to UTF8
  $cur_text =~ s/�/"/g;
  $cur_text =~ s/�/"/g;
  $cur_text =~ s/�/"/g;
  $cur_text =~ s/�/-/g;
  $cur_text =~ s/�/ä/g;
  $cur_text =~ s/�/ö/g;
  $cur_text =~ s/�/ü/g;
  $cur_text =~ s/�/ß/g;
  $cur_text =~ s/�/Ä/g;
  $cur_text =~ s/�/Ö/g;
  $cur_text =~ s/�/Ü/g;
  $cur_text =~ s/�/''/g;

  # Convert the funny links
  $cur_text =~ s/\[\[(.*?)\]\[(.*?)\]\]/\[$1 $2\]/g;

  # We don't need no stinking A HREF-tags
  $cur_text =~ s#<a href="([^>]*)">([^<]*)</a>#\[$1 $2]#gi;
  $cur_text =~ s#<a href=([^>]*)>([^<]*)</a>#\[$1 $2]#gi;

  # I've got no idea what <nop> does. I'll delete it anyway.
  $cur_text =~ s#<nop>##gi;
  $cur_text =~ s#<k>##gi;
 
  # One and two spaces at the beginning of a line has a different meaning
  # in MediaWiki. Strip it!
  $cur_text =~ s#^  ([^ ])#$1#gm;
  $cur_text =~ s#^ ([^ ])#$1#gm;

  # replace <pre> </pre> sections with space at beginning of line
  # the '?' is for not-greedy matching
  $cur_text =~ s#<pre>\n?(.*?)\n?</pre>#my $x=$1; $x=~s@^@ @gms; $x;#egms;

  # Tables 
  #
  # FW table begin: ^|
  # FW table end:   ^[^|]
  # MW table begin: {| class="wikitable
  # MW table end:   |}
  #
  $cur_text =~ s#^(\|.*?)^[^|]#
      my $x = $1; 
      $x =~ s/\|/ || /gm; 
      $x =~ s/^ \|\|/|-\n|/gm; 
      $x =~ s/^\|-\n[\| ]*?$//gms;
      $x =~ s/\|\|\s*$/|/gm; 
      $x =~ s/\| *\*(.*?)\* *\|/! $1 !/gm; 
      $x =~ s/[!|\s]*$//gms; 
      "{| class=\"wikitable\"\n$x\n|}\n\n" #egms;
  #
  # - replace single pipes with double pipes surrounded by spaces
  # - replace space,pipe,pipe at beginning of line w/ pipe,hyphen,neline,pipe
  # - remove empty rows 
  # - clean up double pipes at line ends
  # - convert bolded table cells from pipes to bangs
  # - remove empty cells from end of line

  # trim tailing newlines
  $cur_text =~ s#\s*$#\n\n#s;

  return $cur_text;
}