ydict-pl

#!/usr/bin/env 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 strict;

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

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;

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

sub ydict
{
  my $url = shift;

  #warn "GET $url\n";
  my $html = get($url);
  my $i    = 0;
  $html =~ s{\r}{}g;
  $html =~ s{\n}{ }g;
  my @URL = grep { !m{tab} } ($html =~ m{(http[^<]+docid=\d+)}sg);
  exit if !scalar(@URL);

  print BOLD YELLOW "\nCorrection: $1\n" if $html =~ m{以下為(\w+)在字典};
  my $pronunciation = $1 if $html =~ m{<div class="pronunciation">(.*?)</div>};
  $pronunciation =~ s{</?[^>]+>}{}g;

  #print BOLD YELLOW "\nPronunciation: $pronunciation\n" if $pronunciation;

  my $indent = "  ";
  $html =~ s{.*?(<div class="def clr nobr">)}{\1};              # behead
  $html =~ s{<div class="ft">.*}{};                             # curtail
  $html =~ s{<p class="example">}{$indent  $cyan}g;
  $html =~ s{<p class="interpret">}{ $green}g;
  $html =~ s{<b>}{$bold}g;
  $html =~ s{</b>}{$reset$cyan}g;
  $html =~ s{</p>}{$reset\n}g;
  $html =~ s{<div class="def clr nobr">}{$reset\n$bold$red}g;
  $html =~ s{<div class="list">}{$reset\t}g;
  $html =~ s{<ol>}{$reset\n}g;
  $html =~ s{<li>}{$indent\*}g;
  $html =~ s{\\}{}g;                                            # bug?
  $html =~ s{</?\w+[^>]*>}{}g;                                  # all tags

  print "$html";
  shift @URL;
  map { ydict($_) } (@URL);
}