] => search for on Wikipedia. You can select a national to only search the national Wikipedia"
when "unpedia"
"unpedia => search for on Uncyclopedia"
else
- "search (or: google ) => search google for | wp => search for on Wikipedia | unpedia => search for on Uncyclopedia"
+ "search (or: google ) => search google for | ddg to search DuckDuckGo | wp => search for on Wikipedia | wa => search for on WolframAlpha | unpedia => search for on Uncyclopedia"
end
end
+ def duckduckgo(m, params)
+ what = params[:words].to_s
+ terms = CGI.escape what
+ url = DDG_API_SEARCH + terms
+
+ hits = @bot.config['duckduckgo.hits']
+ first_pars = params[:firstpar] || @bot.config['duckduckgo.first_par']
+ single = params[:lucky] || (hits == 1 and first_pars == 1)
+
+ begin
+ feed = @bot.httputil.get(url)
+ raise unless feed
+ rescue => e
+ m.reply "error duckduckgoing for #{what}"
+ return
+ end
+
+ xml = REXML::Document.new feed
+ heading = xml.elements['//Heading/text()'].to_s
+ # answer is returned for calculations
+ answer = xml.elements['//Answer/text()'].to_s
+ # abstract is returned for definitions etc
+ abstract = xml.elements['//AbstractText/text()'].to_s
+ abfrom = ""
+ unless abstract.empty?
+ absrc = xml.elements['//AbstractSource/text()'].to_s
+ aburl = xml.elements['//AbstractURL/text()'].to_s
+ unless absrc.empty? and aburl.empty?
+ abfrom = " --"
+ abfrom << " " << absrc unless absrc.empty?
+ abfrom << " " << aburl unless aburl.empty?
+ end
+ end
+
+ # but also definition (yes, you can have both, see e.g. printf)
+ definition = xml.elements['//Definition/text()'].to_s
+ deffrom = ""
+ unless definition.empty?
+ defsrc = xml.elements['//Definition/@source/text()'].to_s
+ defurl = xml.elements['//Definition/@url/text()'].to_s
+ unless defsrc.empty? and defurl.empty?
+ deffrom = " --"
+ deffrom << " " << defsrc unless defsrc.empty?
+ deffrom << " " << defurl unless defurl.empty?
+ end
+ end
+
+ if heading.empty? and answer.empty? and abstract.empty? and definition.empty?
+ m.reply "no results"
+ return
+ end
+
+ # if we got a one-shot answer (e.g. a calculation, return it)
+ unless answer.empty?
+ m.reply answer
+ return
+ end
+
+ # otherwise, return the abstract, followed by as many hits as found
+ unless heading.empty? or abstract.empty?
+ m.reply "%{bold}%{heading}:%{bold} %{abstract}%{abfrom}" % {
+ :bold => Bold, :heading => heading,
+ :abstract => abstract, :abfrom => abfrom
+ }
+ end
+ unless heading.empty? or definition.empty?
+ m.reply "%{bold}%{heading}:%{bold} %{abstract}%{abfrom}" % {
+ :bold => Bold, :heading => heading,
+ :abstract => definition, :abfrom => deffrom
+ }
+ end
+ # return zeroclick search results
+ links, texts = [], []
+ xml.elements.each("//Results/Result/FirstURL") { |element|
+ links << element.text
+ break if links.size == hits
+ }
+ return if links.empty?
+
+ xml.elements.each("//Results/Result/Text") { |element|
+ texts << " #{element.text}"
+ break if links.size == hits
+ }
+ # TODO see treatment of `single` in google search
+
+ single ||= (links.length == 1)
+ pretty = []
+ links.each_with_index do |u, i|
+ t = texts[i]
+ pretty.push("%{n}%{b}%{t}%{b}%{sep}%{u}" % {
+ :n => (single ? "" : "#{i}. "),
+ :sep => (single ? " -- " : ": "),
+ :b => Bold, :t => t, :u => u
+ })
+ end
+
+ result_string = pretty.join(" | ")
+
+ # If we return a single, full result, change the output to a more compact representation
+ if single
+ fp = first_pars > 0 ? " -- #{Utils.get_first_pars(links, first_pars)}" : ""
+ m.reply("Result for %{what}: %{string}%{fp}" % {
+ :what => what, :string => result_string, :fp => fp
+ }, :overlong => :truncate)
+ return
+ end
+
+ m.reply "Results for #{what}: #{result_string}", :split_at => /\s+\|\s+/
+
+ return unless first_pars > 0
+
+ Utils.get_first_pars urls, first_pars, :message => m
+ end
+
def google(m, params)
what = params[:words].to_s
if what.match(/^define:/)
@@ -90,48 +215,32 @@ class SearchPlugin < Plugin
m.reply "error googling for #{what}"
return
end
- results = wml.match('').pre_match.scan(GOOGLE_WAP_LINK)
+ results = wml.scan(GOOGLE_WAP_LINK)
if results.length == 0
m.reply "no results found for #{what}"
return
end
- single ||= (results.length==1)
+ results = results.map {|result|
+ url = CGI::parse(Utils.decode_html_entities(result[0]))['q'].first
+ title = Utils.decode_html_entities(result[1].gsub(/<\/?[^>]+>/, ''))
+ [url, title] unless url.empty? or title.empty?
+ }.reject {|item| not item}[0..hits]
- urls = Array.new
- n = 0
- results = results[0...hits].map { |res|
- n += 1
- t = res[2].ircify_html(:img => "[%{src} %{alt} %{dimensions}]").strip
- u = URI.unescape(res[0] || res[1])
- urls.push(u)
- "%{n}%{b}%{t}%{b}%{sep}%{u}" % {
- :n => (single ? "" : "#{n}. "),
- :sep => (single ? " -- " : ": "),
- :b => Bold, :t => t, :u => u
- }
- }
+ result_string = results.map {|url, title| "#{Bold}#{title}#{NormalText}: #{url}"}
if params[:lucky]
- m.reply results.first
+ m.reply result_string.first
+ Utils.get_first_pars([results.map {|url, title| url}.first], first_pars, :message => m)
return
end
- result_string = results.join(" | ")
-
- # If we return a single, full result, change the output to a more compact representation
- if single
- m.reply "Result for %s: %s -- %s" % [what, result_string, Utils.get_first_pars(urls, first_pars)], :overlong => :truncate
- return
- end
-
- m.reply "Results for #{what}: #{result_string}", :split_at => /\s+\|\s+/
+ m.reply "Results for #{what}: #{result_string.join(' | ')}", :split_at => /\s+\|\s+/
return unless first_pars > 0
- Utils.get_first_pars urls, first_pars, :message => m
-
+ Utils.get_first_pars(results.map {|url, title| url}, first_pars, :message => m)
end
def google_define(m, what, params)
@@ -172,7 +281,7 @@ class SearchPlugin < Plugin
searchfor = CGI.escape(what)
debug "Getting gcalc thing: #{searchfor.inspect}"
- url = GOOGLE_SEARCH + searchfor
+ url = GOOGLE_WAP_SEARCH + searchfor
begin
html = @bot.httputil.get(url)
@@ -182,88 +291,90 @@ class SearchPlugin < Plugin
end
debug "#{html.size} bytes of html recieved"
+ debug html
- results = html.scan(GOOGLE_CALC_RESULT)
- debug "results: #{results.inspect}"
+ candidates = html.match(GOOGLE_CALC_RESULT)
+ debug "candidates: #{candidates.inspect}"
- if results.length != 1
+ if candidates.nil?
m.reply "couldn't calculate #{what}"
return
end
+ result = candidates[1].remove_nonascii
- result = results[0][0].ircify_html
debug "replying with: #{result.inspect}"
- m.reply "#{result}"
+ m.reply result.ircify_html
end
- def gcount(m, params)
+ def gdef(m, params)
what = params[:words].to_s
- searchfor = CGI.escape(what)
+ searchfor = CGI.escape("define " + what)
- debug "Getting gcount thing: #{searchfor.inspect}"
- url = GOOGLE_SEARCH + searchfor
+ debug "Getting gdef thing: #{searchfor.inspect}"
+ url = GOOGLE_WAP_SEARCH + searchfor
begin
- html = @bot.httputil.get(url)
+ resp = @bot.httputil.get(url, resp: true)
rescue => e
- m.reply "error googlecounting #{what}"
+ m.reply "error googledefining #{what}"
return
end
- debug "#{html.size} bytes of html recieved"
-
- results = html.scan(GOOGLE_COUNT_RESULT)
- debug "results: #{results.inspect}"
+ results = resp.xpath(GOOGLE_XPATH_DEF).map(&:content)
- if results.length != 1
- m.reply "couldn't count #{what}"
- return
+ if results.empty?
+ m.reply "couldn't find a definition for #{what} on Google"
+ else
+ m.reply "#{results.first} -- #{results[1..-1].join(' ')}"
end
-
- result = results[0][0].ircify_html
- debug "replying with: #{result.inspect}"
- m.reply "total results: #{result}"
-
end
- def gdef(m, params)
+ def wolfram(m, params)
what = params[:words].to_s
- searchfor = CGI.escape("define " + what)
-
- debug "Getting gdef thing: #{searchfor.inspect}"
- url = GOOGLE_WAP_SEARCH + searchfor
+ terms = CGI.escape what
+ url = WOLFRAM_API_SEARCH % {
+ :terms => terms, :key => WOLFRAM_API_KEY
+ }
begin
- html = @bot.httputil.get(url)
+ feed = @bot.httputil.get(url)
+ raise unless feed
rescue => e
- m.reply "error googledefining #{what}"
+ m.reply "error asking WolframAlfa about #{what}"
return
end
+ debug feed
- debug html
- results = html.scan(GOOGLE_DEF_RESULT)
- debug "results: #{results.inspect}"
-
- if results.length != 1
- m.reply "couldn't find a definition for #{what} on Google"
+ xml = REXML::Document.new feed
+ if xml.elements['/queryresult'].attributes['error'] == "true"
+ m.reply xml.elements['/queryresult/error/text()'].to_s
return
end
-
- gdef_link = "http://www.google.com" + CGI.unescapeHTML(results[0][0]) # could be used to extract all defs
- head = results[0][1].ircify_html
- text = results[0][2].ircify_html
- link = results[0][3]
- m.reply "#{head} -- #{link}\n#{text}"
-
- ### gdef_link could be used for something like
- # html_defs = @bot.httputil.get(gdef_link)
- # related_index = html_defs.index(/Related phrases:/, 0)
- # defs_index = html_defs.index(/Definitions of /, related_index)
-
- # related = html_defs[related_index..defs_index]
- # defs = html_defs[defs_index..-1]
-
- # m.reply defs.gsub('
','').ircify_html
+ unless xml.elements['/queryresult'].attributes['success'] == "true"
+ m.reply "no data available"
+ return
+ end
+ answer_type, answer = [], []
+ xml.elements.each("//pod") { |element|
+ answer_type << element.attributes['title']
+ answer << element.elements['subpod/plaintext'].text
+ }
+ # find the first answer that isn't nil,
+ # starting on the second pod in the array
+ n = 1
+ answer[1..-1].each { |a|
+ break unless a.nil?
+ n += 1
+ }
+ if answer[n].nil?
+ m.reply "no results"
+ return
+ end
+ # strip spaces, pipes, and line breaks
+ sep = Bold + ' :: ' + Bold
+ chars = [ [/\n/, sep], [/\t/, " "], [/\s+/, " "], ["|", "-"] ]
+ chars.each { |c| answer[n].gsub!(c[0], c[1]) }
+ m.reply answer_type[n] + sep + answer[n]
end
def wikipedia(m, params)
@@ -286,44 +397,17 @@ class SearchPlugin < Plugin
params[:firstpar] = @bot.config['wikipedia.first_par']
return google(m, params)
end
-
- def gtime(m, params)
- where = params[:words].to_s
- where.sub!(/^\s*in\s*/, '')
- searchfor = CGI.escape("time in " + where)
- url = GOOGLE_SEARCH + searchfor
-
- begin
- html = @bot.httputil.get(url)
- rescue => e
- m.reply "Error googletiming #{where}"
- return
- end
-
- debug html
- results = html.scan(GOOGLE_TIME_RESULT)
- debug "results: #{results.inspect}"
-
- if results.length != 1
- m.reply "Couldn't find the time for #{where} on Google"
- return
- end
-
- time = results[0][0].ircify_html
- m.reply "#{time}"
- end
end
plugin = SearchPlugin.new
+plugin.map "ddg *words", :action => 'duckduckgo', :threaded => true
plugin.map "search *words", :action => 'google', :threaded => true
plugin.map "google *words", :action => 'google', :threaded => true
plugin.map "lucky *words", :action => 'lucky', :threaded => true
-plugin.map "gcount *words", :action => 'gcount', :threaded => true
plugin.map "gcalc *words", :action => 'gcalc', :threaded => true
plugin.map "gdef *words", :action => 'gdef', :threaded => true
-plugin.map "gtime *words", :action => 'gtime', :threaded => true
+plugin.map "wa *words", :action => 'wolfram', :threaded => true
plugin.map "wp :lang *words", :action => 'wikipedia', :requirements => { :lang => /^\w\w\w?$/ }, :threaded => true
plugin.map "wp *words", :action => 'wikipedia', :threaded => true
plugin.map "unpedia *words", :action => 'unpedia', :threaded => true
-