# for most languages
GOOGLE_SEARCH = "http://www.google.com/search?oe=UTF-8&q="
-GOOGLE_WAP_SEARCH = "http://www.google.com/wml/search?hl=en&q="
-GOOGLE_WAP_LINK = /<a accesskey="(\d)" href=".*?u=(.*?)">(.*?)<\/a>/im
-GOOGLE_CALC_RESULT = %r{<p><table><tr><td><img src=/images/calc_img\.gif(?: alt="")?></td><td> </td><td nowrap>(?:<h2 class=r>)?<font size=\+1><b>(.+)</b>(?:</h2>)?</td></tr><tr><td>}
-GOOGLE_DEF_RESULT = %r{<p> (Web definitions for .*?)<br/>(.*?)<br/>(.*?)\s-\s+<a href}
+GOOGLE_WAP_SEARCH = "http://www.google.com/m/search?hl=en&q="
+GOOGLE_WAP_LINK = /result">(?:<div[^>]*>)?<a href="([^"]+)"[^>]*>(.*?)<\/a>/im
+GOOGLE_CALC_RESULT = %r{<img src=/images/calc_img\.gif(?: width=40 height=30 alt="")?>.*?<h[1-6] class=r[^>]*><b>(.+?)</b>}
+GOOGLE_COUNT_RESULT = %r{<font size=-1>Results <b>1<\/b> - <b>10<\/b> of about <b>(.*)<\/b> for}
+GOOGLE_DEF_RESULT = %r{onebox_result">\s*(.*?)\s*<br/>\s*(.*?)<table}
+GOOGLE_TIME_RESULT = %r{alt="Clock"></td><td valign=[^>]+>(.+?)<(br|/td)>}
class SearchPlugin < Plugin
Config.register Config::IntegerValue.new('google.hits',
"gcalc <equation> => use the google calculator to find the answer to <equation>"
when "gdef"
"gdef <term(s)> => use the google define mechanism to find a definition of <term(s)>"
+ when "gtime"
+ "gtime <location> => use the google clock to find the current time at <location>"
when "wp"
"wp [<code>] <string> => search for <string> on Wikipedia. You can select a national <code> to only search the national Wikipedia"
when "unpedia"
def google(m, params)
what = params[:words].to_s
+ if what.match(/^define:/)
+ return google_define(m, what, params)
+ end
+
searchfor = CGI.escape what
# This method is also called by other methods to restrict searching to some sites
if params[:site]
url = GOOGLE_WAP_SEARCH + site + searchfor
hits = params[:hits] || @bot.config['google.hits']
+ hits = 1 if params[:lucky]
first_pars = params[:firstpar] || @bot.config['google.first_par']
- single = (hits == 1 and first_pars == 1)
+ single = params[:lucky] || (hits == 1 and first_pars == 1)
begin
wml = @bot.httputil.get(url)
return
end
results = wml.scan(GOOGLE_WAP_LINK)
+
if results.length == 0
m.reply "no results found for #{what}"
return
end
+
single ||= (results.length==1)
- urls = Array.new
- results = results[0...hits].map { |res|
- n = res[0]
- t = Utils.decode_html_entities res[2].gsub(filter, '').strip
- u = URI.unescape res[1]
- urls.push(u)
- single ? u : "#{n}. #{Bold}#{t}#{Bold}: #{u}"
- }.join(" | ")
+ pretty = []
+
+ begin
+ urls = Array.new
+
+ debug results
+ results.each do |res|
+ t = res[1].ircify_html(:img => "[%{src} %{alt} %{dimensions}]").strip
+ u = res[0]
+ if u.sub!(%r{^http://www.google.com/aclk\?},'')
+ u = CGI::parse(u)['adurl'].first
+ debug "skipping ad for #{u}"
+ next
+ elsif u.sub!(%r{^http://www.google.com/gwt/x\?},'')
+ u = CGI::parse(u)['u'].first
+ elsif u.sub!(%r{^/url\?},'')
+ u = CGI::parse(u)['q'].first
+ end
+ urls.push(u)
+ pretty.push("%{n}%{b}%{t}%{b}%{sep}%{u}" % {
+ :n => (single ? "" : "#{urls.length}. "),
+ :sep => (single ? " -- " : ": "),
+ :b => Bold, :t => t, :u => u
+ })
+ break if urls.length == hits
+ end
+ rescue => e
+ m.reply "failed to understand what google found for #{what}"
+ error e
+ debug wml
+ debug results
+ return
+ end
+
+ if params[:lucky]
+ m.reply pretty.first
+ return
+ end
+
+ result_string = pretty.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, results, Utils.get_first_pars(urls, first_pars)], :overlong => :truncate
+ 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}: #{results}", :split_at => /\s+\|\s+/
+ m.reply "Results for #{what}: #{result_string}", :split_at => /\s+\|\s+/
return unless first_pars > 0
end
+ def google_define(m, what, params)
+ begin
+ wml = @bot.httputil.get(GOOGLE_SEARCH + CGI.escape(what))
+ raise unless wml
+ rescue => e
+ m.reply "error googling for #{what}"
+ return
+ end
+
+ begin
+ related_index = wml.index(/Related phrases:/, 0)
+ raise unless related_index
+ defs_index = wml.index(/Definitions of <b>/, related_index)
+ raise unless defs_index
+ defs_end = wml.index(/<input/, defs_index)
+ raise unless defs_end
+ rescue => e
+ m.reply "no results found for #{what}"
+ return
+ end
+
+ related = wml[related_index...defs_index]
+ defs = wml[defs_index...defs_end]
+
+ m.reply defs.ircify_html(:a_href => Underline), :split_at => (Underline + ' ')
+
+ end
+
+ def lucky(m, params)
+ params.merge!(:lucky => true)
+ google(m, params)
+ end
+
def gcalc(m, params)
what = params[:words].to_s
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)
end
debug "#{html.size} bytes of html recieved"
-
- results = html.scan(GOOGLE_CALC_RESULT)
+ debug html
+
+ candidates = html.match(/font-weight:bold">(.*?)<\/(?:span|div)>/)
+ debug "candidates: #{candidates.inspect}"
+
+ if candidates.nil?
+ m.reply "couldn't calculate #{what}"
+ return
+ end
+ result = candidates[1]
+
+ debug "replying with: #{result.inspect}"
+ m.reply result.ircify_html
+ end
+
+ def gcount(m, params)
+ what = params[:words].to_s
+ searchfor = CGI.escape(what)
+
+ debug "Getting gcount thing: #{searchfor.inspect}"
+ url = GOOGLE_SEARCH + searchfor
+
+ begin
+ html = @bot.httputil.get(url)
+ rescue => e
+ m.reply "error googlecounting #{what}"
+ return
+ end
+
+ debug "#{html.size} bytes of html recieved"
+
+ results = html.scan(GOOGLE_COUNT_RESULT)
debug "results: #{results.inspect}"
-
+
if results.length != 1
- m.reply "couldn't calculate #{what}"
+ m.reply "couldn't count #{what}"
return
end
-
+
result = results[0][0].ircify_html
debug "replying with: #{result.inspect}"
- m.reply "#{result}"
+ m.reply "total results: #{result}"
+
end
def gdef(m, params)
what = params[:words].to_s
searchfor = CGI.escape("define " + what)
-
+
debug "Getting gdef thing: #{searchfor.inspect}"
url = GOOGLE_WAP_SEARCH + searchfor
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"
return
end
-
+
head = results[0][0].ircify_html
text = results[0][1].ircify_html
- link = results[0][2]
- m.reply "#{head} -- #{link}\n#{text}"
+ m.reply "#{head} -- #{text}"
end
def wikipedia(m, params)
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 "search *words", :action => 'google'
-plugin.map "google *words", :action => 'google'
-plugin.map "gcalc *words", :action => 'gcalc'
-plugin.map "gdef *words", :action => 'gdef'
-plugin.map "wp :lang *words", :action => 'wikipedia', :requirements => { :lang => /^\w\w\w?$/ }
-plugin.map "wp *words", :action => 'wikipedia'
-plugin.map "unpedia *words", :action => 'unpedia'
+plugin.map "search *words", :action => 'google', :threaded => true
+plugin.map "google *words", :action => 'google', :threaded => true
+plugin.map "lucky *words", :action => 'lucky', :threaded => true
+plugin.map "gcount *words", :action => 'gcount', :threaded => true
+plugin.map "gcalc *words", :action => 'gcalc', :threaded => true
+plugin.map "gdef *words", :action => 'gdef', :threaded => true
+plugin.map "gtime *words", :action => 'gtime', :threaded => true
+plugin.map "wp :lang *words", :action => 'wikipedia', :requirements => { :lang => /^\w\w\w?$/ }, :threaded => true
+plugin.map "wp *words", :action => 'wikipedia', :threaded => true
+plugin.map "unpedia *words", :action => 'unpedia', :threaded => true
projects / user / henk / code / ruby / rbot.git / blobdiff