simple_quoter

#!/usr/bin/env perl

# This is a short script that tries to automatically
# insert curling quotes.

# By David A. Wheeler
# Copyright the Linux Foundation
# SPDX-License-Identifier: MIT

use utf8;
use open qw( :std :encoding(UTF-8) );
use feature 'unicode_strings';

$in_pre = 0;

while (<>) {
  # Handle markdown ~~~~
  if (/^~~~~/) {
    $in_pre = !$in_pre;
    print;
    next;
  }
  if ($in_pre) {
    print;
    next;
  }

  # Fix curling quotes like “a” and ‘a’

  # Single quotes begin/end a word
  s/^'/‘/;
  s/(<br>|[ \(\[\{])(\**)'/\1\2‘/g;
  s/'$/’/;
  s/'(\**)([ ,.\)\]\}\:])/’\1\2/g;

  # Contractions
  s/n't/n’t/;
  s/([A-Za-z])'s/\1’s/;

  # Double quotes
  # space * "
  s/^(\**)"/\1“/;
  s/(<br>|[ \(\[\{])(\**)"/\1\2“/g;
  s/"(\**)$/”\1/;
  s/"(\**)([ ,.\)\]\}])/”\1\2/g;

  # If they're lonely, convert them back
  s/ [“”] / " /g;
  s/ [‘’] / ' /g;

  # Fix URLs / HTML
  s/["“”](https?:\/\/[^ "'“”‘’]*)["“”]/"\1"/g;
  s/['‘’](https?:\/\/[^ "'“”‘’]*)['‘’]/'\1'/g;
  s/ ([a-z]+)=["“”]([^ "“”]*)["“”]/ \1="\2"/g;
  s/ ([a-z]+)=['‘’]([^ '‘’]*)['‘’]/ \1='\2'/g;
  s/["“”]>/">/g;
  s/['‘’]>/'>/g;

  print;
}