Blame CleanTxt

Packit 504f36
#! /usr/bin/perl -w
Packit 504f36
Packit 504f36
# Script to take the output of nroff -man and remove all the backspacing and
Packit 504f36
# the page footers and the screen commands etc so that it is more usefully
Packit 504f36
# readable online. In fact, in the latest nroff, intermediate footers don't
Packit 504f36
# seem to be generated any more.
Packit 504f36
Packit 504f36
$blankcount = 0;
Packit 504f36
$lastwascut = 0;
Packit 504f36
$firstheader = 1;
Packit 504f36
Packit 504f36
# Input on STDIN; output to STDOUT.
Packit 504f36
Packit 504f36
while (<STDIN>)
Packit 504f36
  {
Packit 504f36
  s/\x1b\[\d+m//g;   # Remove screen controls "ESC [ number m"
Packit 504f36
  s/.\x8//g;         # Remove "char, backspace"
Packit 504f36
Packit 504f36
  # Handle header lines. Retain only the first one we encounter, but remove
Packit 504f36
  # the blank line that follows. Any others (e.g. at end of document) and the
Packit 504f36
  # following blank line are dropped.
Packit 504f36
Packit 504f36
  if (/^PCRE(\w*)\(([13])\)\s+PCRE\1\(\2\)$/)
Packit 504f36
    {
Packit 504f36
    if ($firstheader)
Packit 504f36
      {
Packit 504f36
      $firstheader = 0;
Packit 504f36
      print;
Packit 504f36
      $lastprinted = $_;
Packit 504f36
      $lastwascut = 0;
Packit 504f36
      }
Packit 504f36
    $_=<STDIN>;       # Remove a blank that follows
Packit 504f36
    next;
Packit 504f36
    }
Packit 504f36
Packit 504f36
  # Count runs of empty lines
Packit 504f36
Packit 504f36
  if (/^\s*$/)
Packit 504f36
    {
Packit 504f36
    $blankcount++;
Packit 504f36
    $lastwascut = 0;
Packit 504f36
    next;
Packit 504f36
    }
Packit 504f36
Packit 504f36
  # If a chunk of lines has been cut out (page footer) and the next line
Packit 504f36
  # has a different indentation, put back one blank line.
Packit 504f36
Packit 504f36
  if ($lastwascut && $blankcount < 1 && defined($lastprinted))
Packit 504f36
    {
Packit 504f36
    ($a) = $lastprinted =~ /^(\s*)/;
Packit 504f36
    ($b) = $_ =~ /^(\s*)/;
Packit 504f36
    $blankcount++ if ($a ne $b);
Packit 504f36
    }
Packit 504f36
Packit 504f36
  # We get here only when we have a non-blank line in hand. If it was preceded
Packit 504f36
  # by 3 or more blank lines, read the next 3 lines and see if they are blank.
Packit 504f36
  # If so, remove all 7 lines, and remember that we have just done a cut.
Packit 504f36
Packit 504f36
  if ($blankcount >= 3)
Packit 504f36
    {
Packit 504f36
    for ($i = 0; $i < 3; $i++)
Packit 504f36
      {
Packit 504f36
      $next[$i] = <STDIN>;
Packit 504f36
      $next[$i] = "" if !defined $next[$i];
Packit 504f36
      $next[$i] =~ s/\x1b\[\d+m//g;   # Remove screen controls "ESC [ number m"
Packit 504f36
      $next[$i] =~ s/.\x8//g;         # Remove "char, backspace"
Packit 504f36
      }
Packit 504f36
Packit 504f36
    # Cut out chunks of the form <3 blanks><non-blank><3 blanks>
Packit 504f36
Packit 504f36
    if ($next[0] =~ /^\s*$/ &&
Packit 504f36
        $next[1] =~ /^\s*$/ &&
Packit 504f36
        $next[2] =~ /^\s*$/)
Packit 504f36
      {
Packit 504f36
      $blankcount -= 3;
Packit 504f36
      $lastwascut = 1;
Packit 504f36
      }
Packit 504f36
Packit 504f36
    # Otherwise output the saved blanks, the current, and the next three
Packit 504f36
    # lines. Remember the last printed line.
Packit 504f36
Packit 504f36
    else
Packit 504f36
      {
Packit 504f36
      for ($i = 0; $i < $blankcount; $i++) { print "\n"; }
Packit 504f36
      print;
Packit 504f36
      for ($i = 0; $i < 3; $i++)
Packit 504f36
        {
Packit 504f36
        $next[$i] =~ s/.\x8//g;
Packit 504f36
        print $next[$i];
Packit 504f36
        $lastprinted = $_;
Packit 504f36
        }
Packit 504f36
      $lastwascut = 0;
Packit 504f36
      $blankcount = 0;
Packit 504f36
      }
Packit 504f36
    }
Packit 504f36
Packit 504f36
  # This non-blank line is not preceded by 3 or more blank lines. Output
Packit 504f36
  # any blanks there are, and the line. Remember it. Force two blank lines
Packit 504f36
  # before headings.
Packit 504f36
Packit 504f36
  else
Packit 504f36
    {
Packit 504f36
    $blankcount = 2 if /^\S/ && !/^Last updated/ && !/^Copyright/ &&
Packit 504f36
      defined($lastprinted);
Packit 504f36
    for ($i = 0; $i < $blankcount; $i++) { print "\n"; }
Packit 504f36
    print;
Packit 504f36
    $lastprinted = $_;
Packit 504f36
    $lastwascut = 0;
Packit 504f36
    $blankcount = 0;
Packit 504f36
    }
Packit 504f36
  }
Packit 504f36
Packit 504f36
# End