doc/doc-docbook/Tidytxt

   1 #! /usr/bin/perl
   2
   3 # $Cambridge: exim/doc/doc-docbook/Tidytxt,v 1.2 2006/02/01 11:01:01 ph10 Exp $
   4
   5 # Script to tidy up the output of w3m when it makes a text file. First we
   6 # convert sequences of blank lines into a single blank line, to get everything
   7 # uniform. Then we go through and insert blank lines before chapter and
   8 # sections, also converting chapter titles to uppercase.
   9
  10 @lines = <>;
  11
  12 $lastwasblank = 0;
  13 foreach $line (@lines)
  14   {
  15   if ($line =~ /^\s*$/)
  16     {
  17     $line = "" if $lastwasblank;
  18     $lastwasblank = 1;
  19     next;
  20     }
  21   $lastwasblank = 0;
  22   }
  23
  24 # Find start of TOC, uppercasing its title
  25
  26 for ($i = 0; $i < scalar @lines; $i++)
  27   {
  28   $lines[$i] = "TABLE OF CONTENTS\n" if $lines[$i] =~ /^Table of Contents/;
  29   last if $lines[$i] =~ /^1. /;
  30   }
  31
  32 # Find start of first chapter
  33
  34 for ($i++; $i < scalar @lines; $i++)
  35   { last if $lines[$i] =~ /^1. /; }
  36
  37 # Process the body. We can detect the starts of chapters and sections by
  38 # looking for preceding and following blank lines, and then matching against
  39 # the numbers.
  40
  41 $chapter = 0;
  42 for (; $i < scalar @lines; $i++)
  43   {
  44   next if $lines[$i-1] !~ /^$/ || $lines[$i+1] !~ /^$/;
  45
  46   # Start of chapter
  47
  48   if ($lines[$i] =~ /^(\d+)\. / && $1 == $chapter + 1)
  49     {
  50     $chapter++;
  51     $section = 0;
  52     $lines[$i] = "\n\n" . ("=" x 79) . "\n" . uc($lines[$i]);
  53     }
  54
  55   # Start of next section
  56
  57   elsif ($lines[$i] =~ /^(\d+)\.(\d+) / && $1 == $chapter && $2 == $section + 1)
  58     {
  59     $section++;
  60     $lines[$i] = "\n$lines[$i]" . "-" x (length($lines[$i]) - 1) . "\n";
  61     }
  62   }
  63
  64 print @lines;
  65
  66 # End