X-Git-Url: https://git.netwichtig.de/gitweb/?a=blobdiff_plain;f=data%2Frbot%2Fplugins%2Ftranslator.rb;h=833701c36cb3eee19f8bef920a492811b9039a66;hb=633fdf34ed152f06b123e7b3167177bf385f49bb;hp=8617c321e8196a2f86bebb69d3eba93893f62ede;hpb=c3e2d59d3e081141fcdcd5ee43469c3c4a7586db;p=user%2Fhenk%2Fcode%2Fruby%2Frbot.git diff --git a/data/rbot/plugins/translator.rb b/data/rbot/plugins/translator.rb index 8617c321..833701c3 100644 --- a/data/rbot/plugins/translator.rb +++ b/data/rbot/plugins/translator.rb @@ -39,6 +39,14 @@ class Translator @cache = cache end + # Many translators use Mechanize, which changed namespace around version 1.0 + # To support both pre-1.0 and post-1.0 namespaces, we use these auxiliary + # method. The translator still needs to require 'mechanize' on initialization + # if it needs it. + def mechanize + return Mechanize if defined? Mechanize + return WWW::Mechanize + end # whether the translator supports this direction def support?(from, to) @@ -103,84 +111,9 @@ class Translator end end - -class NiftyTranslator < Translator - INFO = '@nifty Translation ' - - def initialize(cache={}) - require 'mechanize' - super(Translator::Direction.all_from_to(%w[ja en zh_CN ko], %w[ja]), cache) - @form = WWW::Mechanize.new. - get('http://nifty.amikai.com/amitext/indexUTF8.jsp'). - forms_with(:name => 'translateForm').last - end - - def do_translate(text, from, to) - @radio = @form.radiobuttons_with(:name => 'langpair').first - @radio.value = "#{from},#{to}".upcase - @radio.check - @form.fields_with(:name => 'sourceText').last.value = text - - @form.submit(@form.buttons_with(:name => 'translate').last). - forms_with(:name => 'translateForm').last.fields_with(:name => 'translatedText').last.value - end -end - - -class ExciteTranslator < Translator - INFO = 'Excite.jp Translation ' - - def initialize(cache={}) - require 'mechanize' - require 'iconv' - - super(Translator::Direction.all_from_to(%w[ja en zh_CN zh_TW ko], %w[ja]), cache) - - @forms = Hash.new do |h, k| - case k - when 'en' - h[k] = open_form('english') - when 'zh_CN', 'zh_TW' - # this way we don't need to fetch the same page twice - h['zh_CN'] = h['zh_TW'] = open_form('chinese') - when 'ko' - h[k] = open_form('korean') - end - end - end - - def open_form(name) - WWW::Mechanize.new.get("http://www.excite.co.jp/world/#{name}"). - forms_with(:name => 'world').first - end - - def do_translate(text, from, to) - non_ja_language = from != 'ja' ? from : to - form = @forms[non_ja_language] - - if non_ja_language =~ /zh_(CN|TW)/ - form_with_fields(:name => 'wb_lp').first.value = "#{from}#{to}".sub(/_(?:CN|TW)/, '').upcase - form_with_fields(:name => 'big5').first.value = ($1 == 'TW' ? 'yes' : 'no') - else - # the en<->ja page is in Shift_JIS while other pages are UTF-8 - text = Iconv.iconv('Shift_JIS', 'UTF-8', text) if non_ja_language == 'en' - form.fields_with(:name => 'wb_lp').first.value = "#{from}#{to}".upcase - end - form.fields_with(:name => 'before').first.value = text - result = form.submit.forms_with(:name => 'world').first.fields_with(:name => 'after').first.value - # the en<->ja page is in Shift_JIS while other pages are UTF-8 - if non_ja_language == 'en' - Iconv.iconv('UTF-8', 'Shift_JIS', result) - else - result - end - - end -end - - class GoogleTranslator < Translator INFO = 'Google Translate ' + URL = 'https://translate.google.com/' LANGUAGES = %w[af sq am ar hy az eu be bn bh bg my ca chr zh zh_CN zh_TW hr @@ -189,71 +122,46 @@ class GoogleTranslator < Translator pt_PT pa ro ru sa sr sd si sk sl es sw sv tg ta tl te th bo tr uk ur uz ug vi cy yi auto] def initialize(cache={}) - require "uri" - require "json" - super(Translator::Direction.all_to_all(LANGUAGES), cache) - end - - def do_translate(text, from, to) - langpair = [from == 'auto' ? '' : from, to].map { |e| e.tr('_', '-') }.join("|") - raw_json = Irc::Utils.bot.httputil.get_response(URI.escape( - "http://ajax.googleapis.com/ajax/services/language/translate?v=1.0&q=#{text}&langpair=#{langpair}")).body - response = JSON.parse(raw_json) - - if response["responseStatus"] != 200 - raise Translator::NoTranslationError, response["responseDetails"] - else - response["responseData"]["translatedText"] - end - end -end - - -class BabelfishTranslator < Translator - INFO = 'AltaVista Babel Fish Translation ' - - def initialize(cache) require 'mechanize' - - @form = WWW::Mechanize.new.get('http://babelfish.altavista.com/babelfish/'). - forms_with(:name => 'frmTrText').first - @lang_list = @form.fields_with(:name => 'lp').first - language_pairs = @lang_list.options.map {|o| o.value.split('_')}. - reject {|p| p.empty?} - super(Translator::Direction.pairs(language_pairs), cache) + super(Translator::Direction.all_to_all(LANGUAGES), cache) end def do_translate(text, from, to) - if @form.fields_with(:name => 'trtext').empty? - @form.add_field!('trtext', text) - else - @form.fields_with(:name => 'trtext').first.value = text - end - @lang_list.value = "#{from}_#{to}" - @form.submit.parser.search("div[@id='result']/div[@style]").inner_html + agent = Mechanize.new + agent.user_agent_alias = 'Linux Mozilla' + page = agent.get URL + form = page.form_with(:id => 'gt-form') + form.sl = from + form.tl = to + form.text = text + page = form.submit + return page.search('#result_box span').first.content end end -class WorldlingoTranslator < Translator - INFO = 'WorldLingo Free Online Translator ' +class YandexTranslator < Translator + INFO = 'Yandex Translator ' + LANGUAGES = %w{ar az be bg ca cs da de el en es et fi fr he hr hu hy it ka lt lv mk nl no pl pt ro ru sk sl sq sr sv tr uk} - LANGUAGES = %w[en fr de it pt es ru nl el sv ar ja ko zh_CN zh_TW] + URL = 'https://translate.yandex.net/api/v1.5/tr.json/translate?key=%s&lang=%s-%s&text=%s' + KEY = 'trnsl.1.1.20140326T031210Z.1e298c8adb4058ed.d93278fea8d79e0a0ba76b6ab4bfbf6ac43ada72' def initialize(cache) require 'uri' + require 'json' super(Translator::Direction.all_to_all(LANGUAGES), cache) end def translate(text, from, to) - response = Irc::Utils.bot.httputil.get_response(URI.escape( - "http://www.worldlingo.com/SEfpX0LV2xIxsIIELJ,2E5nOlz5RArCY,/texttranslate?wl_srcenc=utf-8&wl_trgenc=utf-8&wl_text=#{text}&wl_srclang=#{from.upcase}&wl_trglang=#{to.upcase}")) - # WorldLingo seems to respond an XML when error occurs - case response['Content-Type'] - when %r'text/plain' - response.body - else + res = Irc::Utils.bot.httputil.get_response(URL % [KEY, from, to, URI.escape(text)]) + res = JSON.parse(res.body) + + if res['code'] != 200 raise Translator::NoTranslationError + else + res['text'].join(' ') end end + end class TranslatorPlugin < Plugin @@ -265,27 +173,19 @@ class TranslatorPlugin < Plugin :desc => _("Default destination language to be used with translate command")) TRANSLATORS = { - 'nifty' => NiftyTranslator, - 'excite' => ExciteTranslator, 'google_translate' => GoogleTranslator, - 'babelfish' => BabelfishTranslator, - 'worldlingo' => WorldlingoTranslator, + 'yandex' => YandexTranslator, } def initialize super - + @failed_translators = [] @translators = {} TRANSLATORS.each_pair do |name, c| - begin + watch_for_fail(name) do @translators[name] = c.new(@registry.sub_registry(name)) map "#{name} :from :to *phrase", :action => :cmd_translate, :thread => true - rescue Exception - warning _("Translator %{name} cannot be used: %{reason}") % - {:name => name, :reason => $!} - map "#{name} [*args]", :action => :failed_translator, - :defaults => {:name => name, :reason => $!} end end @@ -297,25 +197,64 @@ class TranslatorPlugin < Plugin update_default end + def watch_for_fail(name, &block) + begin + yield + rescue Exception + debug 'Translator error: '+$!.to_s + debug $@.join("\n") + @failed_translators << { :name => name, :reason => $!.to_s } + + warning _("Translator %{name} cannot be used: %{reason}") % + {:name => name, :reason => $!} + map "#{name} [*args]", :action => :failed_translator, + :defaults => {:name => name, :reason => $!} + end + end + def failed_translator(m, params) m.reply _("Translator %{name} cannot be used: %{reason}") % {:name => params[:name], :reason => params[:reason]} end def help(plugin, topic=nil) - if @translators.has_key?(plugin) - translator = @translators[plugin] - _('%{translator} => Look up phrase using %{info}, supported from -> to languages: %{directions}') % { - :translator => plugin, - :info => translator.class::INFO, - :directions => translator.directions.map do |source, targets| - _('%{source} -> %{targets}') % - {:source => source, :targets => targets.to_a.join(', ')} - end.join(' | ') - } + case (topic.intern rescue nil) + when :failed + unless @failed_translators.empty? + failed_list = @failed_translators.map { |t| _("%{bold}%{translator}%{bold}: %{reason}") % { + :translator => t[:name], + :reason => t[:reason], + :bold => Bold + }} + + _("Failed translators: %{list}") % { :list => failed_list.join(", ") } + else + _("None of the translators failed") + end else - _('Command: , where is one of: %{translators}. If "translator" is used in place of the translator name, the first translator in translator.default_list which supports the specified direction will be picked automatically. Use "help " to look up supported from and to languages') % - {:translators => @translators.keys.join(', ')} + if @translators.has_key?(plugin) + translator = @translators[plugin] + _('%{translator} => Look up phrase using %{info}, supported from -> to languages: %{directions}') % { + :translator => plugin, + :info => translator.class::INFO, + :directions => translator.directions.map do |source, targets| + _('%{source} -> %{targets}') % + {:source => source, :targets => targets.to_a.join(', ')} + end.join(' | ') + } + else + help_str = _('Command: , where is one of: %{translators}. If "translator" is used in place of the translator name, the first translator in translator.default_list which supports the specified direction will be picked automatically. Use "help " to look up supported from and to languages') % + {:translators => @translators.keys.join(', ')} + + help_str << "\n" + _("%{bold}Note%{bold}: %{failed_amt} translators failed, see %{reverse}%{prefix}help translate failed%{reverse} for details") % { + :failed_amt => @failed_translators.size, + :bold => Bold, + :reverse => Reverse, + :prefix => @bot.config['core.address_prefix'].first + } + + help_str + end end end @@ -333,9 +272,22 @@ class TranslatorPlugin < Plugin translator = @default_translators.find {|t| @translators[t].support?(params[:from], params[:to])} if translator - cmd_translate m, params.merge({:translator => translator, :show_provider => true}) + cmd_translate m, params.merge({:translator => translator, :show_provider => false}) else - m.reply _('None of the default translators (translator.default_list) supports translating from %{source} to %{target}') % {:source => params[:from], :target => params[:to]} + # When translate command is used without source language, "auto" as source + # language is assumed. It means that google translator is used and we let google + # figure out what the source language is. + # + # Problem is that the google translator will fail if the system that the bot is + # running on does not have the json gem installed. + if params[:from] == 'auto' + m.reply _("Unable to auto-detect source language due to broken google translator, see %{reverse}%{prefix}help translate failed%{reverse} for details") % { + :reverse => Reverse, + :prefix => @bot.config['core.address_prefix'].first + } + else + m.reply _('None of the default translators (translator.default_list) supports translating from %{source} to %{target}') % {:source => params[:from], :target => params[:to]} + end end end @@ -345,35 +297,56 @@ class TranslatorPlugin < Plugin translator = @translators[tname] from, to, phrase = params[:from], params[:to], params[:phrase].to_s if translator - begin - translation = Timeout.timeout(@bot.config['translator.timeout']) do - translator.translate(phrase, from, to) + watch_for_fail(tname) do + begin + translation = Timeout.timeout(@bot.config['translator.timeout']) do + translator.translate(phrase, from, to) + end + m.reply(if params[:show_provider] + _('%{translation} (provided by %{translator})') % + {:translation => translation, :translator => tname.gsub("_", " ")} + else + translation + end) + + rescue Translator::UnsupportedDirectionError + m.reply _("%{translator} doesn't support translating from %{source} to %{target}") % + {:translator => tname, :source => from, :target => to} + rescue Translator::NoTranslationError + m.reply _('%{translator} failed to provide a translation') % + {:translator => tname} + rescue Timeout::Error + m.reply _('The translator timed out') end - m.reply(if params[:show_provider] - _('%{translation} (provided by %{translator})') % - {:translation => translation, :translator => tname.gsub("_", " ")} - else - translation - end) - - rescue Translator::UnsupportedDirectionError - m.reply _("%{translator} doesn't support translating from %{source} to %{target}") % - {:translator => tname, :source => from, :target => to} - rescue Translator::NoTranslationError - m.reply _('%{translator} failed to provide a translation') % - {:translator => tname} - rescue Timeout::Error - m.reply _('The translator timed out') end else m.reply _('No translator called %{name}') % {:name => tname} end end + + # URL translation has nothing to do with Translators so let's make it + # separate, and Google exclusive for now + def cmd_translate_url(m, params) + params[:to] = @bot.config['translator.destination'] if params[:to].nil? + params[:from] ||= 'auto' + + translate_url = "http://translate.google.com/translate?sl=%{from}&tl=%{to}&u=%{url}" % { + :from => params[:from], + :to => params[:to], + :url => CGI.escape(params[:url].to_s) + } + + m.reply(translate_url) + end end plugin = TranslatorPlugin.new req = Hash[*%w(from to).map { |e| [e.to_sym, /#{plugin.languages.join("|")}/] }.flatten] +plugin.map 'translate [:from] [:to] :url', + :action => :cmd_translate_url, :requirements => req.merge(:url => %r{^https?://[^\s]*}) +plugin.map 'translator [:from] [:to] :url', + :action => :cmd_translate_url, :requirements => req.merge(:url => %r{^https?://[^\s]*}) plugin.map 'translate [:from] [:to] *phrase', :action => :cmd_translator, :thread => true, :requirements => req plugin.map 'translator [:from] [:to] *phrase',