#!/usr/bin/perl
# Yen-Ming Lee <leeym@leeym.com>
# 2010/06/01
#
# ----------------------------------------------------------------------------
# "THE PEARL-TEA-WARE LICENSE", based on "THE BEER-WARE LICENSE":
# <leeym@leeym.com> wrote this file. As long as you retain this notice you
# can do whatever you want with this stuff. If we meet some day, and you think
# this stuff is worth it, you can buy me a pearl tea in return. Yen-Ming Lee
# ----------------------------------------------------------------------------

use FindBin qw($Script);
use Term::ANSIColor qw(:constants);
use LWP::Simple;
use URI::Escape;
use Data::Dumper;
use Getopt::Std;
use strict;

use utf8;
binmode(STDOUT, ':encoding(utf8)');

my %opt;
getopts('vd', \%opt);

my $word = shift;
die "usage: $Script word\n" if !$word;

my $red    = RED;
my $bold   = BOLD;
my $cyan   = CYAN;
my $reset  = RESET;
my $green  = GREEN;
my $yellow = YELLOW;
my %URL;

my $base = 'http://tw.dictionary.search.yahoo.com';
ydict("$base/search?fr=ydict&p=" . uri_escape($word));

sub ydict
{
  my $url = shift;
  warn "GET $url\n" if $opt{d};
  my $html = get($url);
  $html =~ s{\r}{}g;
  $html =~ s{\n}{ }g;
  my @URL = grep { !m{(tab|\*)} } ($html =~ m{(http[^"]+docid=[^"]+)}sg);

  if ($html =~ m{<a href="([^"]+)"><i>(\w+)</i>})
  {
    my ($url2, $correction) = ($1, $2);
    print "$bold$yellow\nCorrection: $correction$reset\n";
    return ydict("$base$url2");
  }
  if ($html =~ m{isZrp: true})
  {
    print "Not found\n";
    exit;
  }
  my $pronunciation = $1 if $html =~ m{<div class="pronunciation">(.*?)</div>};
  $pronunciation =~ s{</?[^>]+>}{}g;

  print "\n${bold}{$yellow}Pronunciation: $pronunciation${reset}\n"
    if $pronunciation;

  my $indent = "  ";
  $html =~ s{.*?<li class="explanation_pos_wrapper">}{};    # behead
  $html =~ s{</ol></div>.*}{};                              # curtail
  $html =~ s{</li></ul></li>.*}{};                          # curtail
  $html =~ s{<ul class="extra_wrapper">.*}{};               # curtail
  $html =~ s{<b>}{$bold}g;
  $html =~ s{</span><span class="pos_desc">}{ }g;
  $html =~ s{</b>}{$reset$cyan}g;
  $html =~ s{<samp[^>]*>}{\n$indent  $cyan}g;
  $html =~ s{<p class="explanation">}{$reset\n  * $green}g;
  $html =~ s{<h5 class="explanation_group_hd">}{$reset\n$bold$red}g;
  $html =~ s{</p>}{$reset}g;
  $html =~ s{&#39;}{'}g;
  $html =~ s{\\}{}g;                                                  # bug?
  $html =~ s{</?\w+[^>]*>}{}g;                                        # all tags

  print "$html\n";

  #shift @URL;
  #map { ydict($_) unless $URL{$_}++ } (@URL);
}