#!/usr/bin/perl use File::stat; use strict; if(@ARGV == 0){ process_stream(\*STDIN,\*STDOUT); exit(0); } if($ARGV[0] eq '-h' || @ARGV > 2){ #The one and only command line option. print("Converts gemtext to HTML.\n gmi2html.pl \n\n"); print("If used with no arguments, defaults to converting stdin to stdout.\n\n"); print("If used with one argument, which must be a gmi file, will write to a corresponding html file.\n"); print("If used with two arguments, will convert gemtext in the first file to HTML in the second.\n\n"); print("Links to .gmi files will be changed to .html unless they use the gemini:// protocol descriptor.\n"); print("Links to image files will be converted to img tags.\n"); exit(0); } my $infile = $ARGV[0]; if(! -f $infile){ print STDERR "Input file $infile not found.\n"; exit(1); } my $outfile; if(@ARGV == 1){ #Single file mode. if(!($infile =~ m/\.gmi$/i)){ print STDERR "Input file must end with extension .gmi (run with -h for more information.\n"; exit(1); } $outfile = $infile; $outfile =~ s/\.gmi$/\.html/i; } if(@ARGV == 2){ $outfile=$ARGV[1]; } if(@ARGV != 0){ if(!($outfile =~ m/\.html?$/i)){ print STDERR "Output file must end with extention .html or .htm (This is to prevent accidential overwriting).\n"; exit(1); } } print STDERR "$infile -> $outfile\n"; open(my $infh, "<:encoding(utf-8)", $infile) || die "Failed to open input file $infile"; open(my $outfh, ">:encoding(utf-8)", $outfile) || die "Failed to open input file $outfile"; #print STDERR "Processing.\n"; process_stream($infh, $outfh); close($infh); close($outfh); sub process_stream($$){ my $instream=$_[0]; my $outstream=$_[1]; my @htmlout; #The only reason this can't be done stream-to-stream is the title tag! my $ul=0; #Are we in a bulleted list? my $bc=0; #Are we in a blockquote? my $title; push(@htmlout, "\n\n\n"); while(<$instream>){ s/[\r\n]//g; s/\&/\&/g; s//\>/g; if($_ eq '```'){ #Blockquote toggle if($bc){ push(@htmlout, "\n"); }else{ push(@htmlout, "
\n");
      }
      $bc=!$bc;
      next;
    }
    if($bc){
      push(@htmlout, "$_\n");
      next;
    }
    if($_ =~ m/^\w*$/){ #Empty lines are meaningless in HTML.
      next;
    }
    if(!$ul && substr($_, 0, 2) eq '* '){ #Handling for unordered lists.
      push(@htmlout, "
    \n");
      $ul=1;
    }
    if($ul && !(substr($_, 0, 2) eq '* ')){
      push(@htmlout, "
\n");
      $ul=0;    
    }    
    if($ul){
      $_=substr($_, 2);
      push(@htmlout, "
  • $_
    • \n");
      next;
    }
    if(substr($_, 0, 4) eq '### '){
      push(@htmlout, "
    ".substr($_, 4)."
    \n");
    }elsif(substr($_, 0, 3) eq '## '){
      push(@htmlout, "
    ".substr($_, 3)."
    \n");
    }elsif(substr($_, 0, 2) eq '# '){
      push(@htmlout, "
    ".substr($_, 2)."
    \n");
      if(!$title){
        $title=substr($_, 2);
      }
    }elsif(substr($_, 0, 6) eq '=> '){
      my $link=substr($_, 6);
      my $space=index($link, ' ');
      #$link =~ m/([^ ]+) +(.+)/;
      my $a=substr($link, 0, $space);
      my $b=substr($link, $space+1);
      my $isimage=0;
        if( $a =~ m/\.jpg$/i ||
            $a =~ m/\.png$/i ||
            $a =~ m/\.webp$/i ||
            $a =~ m/\.gif$/i ||
	    $a =~ m/\.svg$/i ||
            $a =~ m/\.avif$/i) {
          $isimage=1;
        }
      if($isimage){
	push(@htmlout, '');
        push(@htmlout, ''.$b.
    \n");
	push(@htmlout, '
    ');
      }else{
	if(substr($a, 0, 9) ne 'gemini://'){
	  $a =~ s/\.gmi$/\.html/i;
	}
        push(@htmlout, '$b
    \n");
      }
    }else{
      push(@htmlout, "
    $_
    \n");
    }
  }
  print $outstream "\n\n\n";
  if($title){
  print $outstream "$title\n";
  }
  print $outstream '';
  print $outstream ''; 
  print $outstream "
    ";
  while(my $line=shift(@htmlout)){
    print $outstream $line;
  }
  print $outstream "
    \n\n";
}